v0.4.7

What's Changed

• Eval all cache arrays before moving streams by @lucasnewman in #862
• Fix MLX server thread ownership for model loading by @steipete in #863
• Update transformers dependency version to 5.14.0 by @Blaizzy in #864
• Bump version to 0.4.7 by @Blaizzy in #866
• Add a registry to classify supported audio models by kind by @Lazarus-931 in #865
• Add arktts (Audio8-TTS-Preview-0.6b): DualAR TTS with zero-shot voice cloning by @xocialize in #867
• fix(dsp): BS.1770 integrated loudness misses the 997 Hz calibration anchor by @gaoflow in #860
• Add Chatterbox Multilingual v3 model by @lucasnewman in #869

New Contributors

• @steipete made their first contribution in #863
• @gaoflow made their first contribution in #860

Full Changelog: v0.4.6...v0.4.7