Charles Hasseandcapy-ai-staging[bot] ad5d64c0a3 feat(backend): run Chatterbox multilingual on MLX for Apple Silicon ...

The Chatterbox backend is pinned to the CPU on macOS, so voice cloning runs at
roughly 4x realtime there. This adds an MLX/Metal backend for the same engine and
selects it on Apple Silicon, mirroring the split the qwen engine already makes
between mlx_backend and pytorch_backend.

Measured on an M4 Max (36 GB) with a cloned pt-BR profile, same API, same profile,
model already loaded:

  short sentence (1.8s of audio):  7.4-9.2s  ->  1.2s
  longer sentence (5.0s of audio): 23.7s     ->  3.1s

The model config for chatterbox-tts is now backend aware, same as the qwen configs,
so the download matches the backend that will consume it.

Nothing changes off Apple Silicon: the PyTorch backend is still selected there, and
the CPU pinning it relies on is untouched.

2026-10-04 00:25:58 +00:00

..

__init__.py

feat(backend): run Chatterbox multilingual on MLX for Apple Silicon

2026-10-04 00:25:58 +00:00

base.py

fix(backend): drop the prompt cache before the allocator-empty; note clear_cache is thread-agnostic

2026-10-04 00:25:53 +00:00

chatterbox_backend.py

Merge pull request #1130 from jamiepine/prep/pr-1112

2026-10-04 00:05:46 +00:00

chatterbox_mlx_backend.py

feat(backend): run Chatterbox multilingual on MLX for Apple Silicon

2026-10-04 00:25:58 +00:00

chatterbox_turbo_backend.py

Merge pull request #1130 from jamiepine/prep/pr-1112

2026-10-04 00:05:46 +00:00

hume_backend.py

fix(offline-guard): make every cache check cover the files the forced-offline load reads

2026-10-03 09:35:49 +00:00

kokoro_backend.py

fix(kokoro): trim edges only, keep inter-segment gaps

2026-10-04 00:01:24 +00:00

luxtts_backend.py

fix(backend): prevent unbounded memory accumulation over consecutive TTS generations (#923)

2026-10-04 00:01:18 +00:00

mlx_backend.py

fix(mlx): release the model binding inside the sync bodies so a failing op's traceback cannot pin it

2026-10-04 00:25:53 +00:00

pytorch_backend.py

refactor(stt): name the Whisper sample rate and window constants

2026-10-04 00:01:34 +00:00

qwen_custom_voice_backend.py

fix(backend): prevent unbounded memory accumulation over consecutive TTS generations (#923)

2026-10-04 00:01:18 +00:00

qwen_llm_backend.py

fix(mlx): release the model binding inside the sync bodies so a failing op's traceback cannot pin it

2026-10-04 00:25:53 +00:00