v0.0.4

What's Changed

• Add CSM (Conversational Speech Model) section to README.md by @ivanfioravanti in #56
• Use MLX-based SNAC vocoder for Orpheus by @lucasnewman in #62
• Add Descript neural audio codec by @lucasnewman in #57
• Vectorize overlap-add operation in istft by @lucasnewman in #66
• Align Orpheus sampling parameters with the reference implementation by @lucasnewman in #68
• Improve Korkoro generation performance by @lucasnewman in #73
• Add Speech to Speech Tab to the UI by @freddyaboulton in #79
• Add support for fp16 variant of Sesame by @lucasnewman in #78
• Add (partially-working) voice matching support for Orpheus by @lucasnewman in #75
• Bible Audiobook example by @andrepadez in #43
• removed push notification to ntfy by @andrepadez in #81
• Bump version by @Blaizzy in #82

New Contributors

• @freddyaboulton made their first contribution in #79

Full Changelog: v0.0.3...v0.0.4

v0.0.3

What's Changed

• Add verbose logging and model selection support by @ivanfioravanti in #22
• Pulsating effect by @ivanfioravanti in #23
• Compile the decoder for Kokoro by @lucasnewman in #24
• Play audio segments as they are generated by @lucasnewman in #26
• Evaluate the computation graph before returning results by @ivanfioravanti in #35
• Add Mimi neural audio codec by @lucasnewman in #34
• Add model for Sesame TTS by @lucasnewman in #36
• Sphere speed up during audio generation by @ivanfioravanti in #40
• Added more Voices by @andrepadez in #37
• Feature: External API for Audiobook Generation by @sergenes in #19
• Add EnCodec neural audio codec by @lucasnewman in #46
• Add Suno bark by @Blaizzy in #45
• Update README.md to fix lang_code error by @zboyles in #49
• fix model config by @Blaizzy in #50
• Add Vocos neural audio codec by @lucasnewman in #48
• Fix Kokoro audio generation by @lucasnewman in #52
• Add orpheus by @Blaizzy in #47
• Resample and Transcribe by @chigkim in #51
• Fix vocos config loading by @Blaizzy in #53

New Contributors

• @andrepadez made their first contribution in #37
• @sergenes made their first contribution in #19
• @zboyles made their first contribution in #49
• @chigkim made their first contribution in #51

Full Changelog: v0.0.2...v0.0.3

v0.0.2

What's Changed

• fix workflows and readme by @Blaizzy in #5
• Add soundfile to requirements and Quick Start in README by @ivanfioravanti in #8
• Remove librosa dependency by @lucasnewman in #11
• Add support for command-line playback with the --play argument by @lucasnewman in #10
• Allow receiving text input from stdin or an entry prompt by @lucasnewman in #12
• Add web server and improve audio player by @ivanfioravanti in #14
• Use phonemizer-fork to avoid espeak errors by @rampadc in #17

New Contributors

• @Blaizzy made their first contribution in #5
• @ivanfioravanti made their first contribution in #8
• @lucasnewman made their first contribution in #11
• @rampadc made their first contribution in #17

Full Changelog: v0.0.1...v0.0.2

v0.0.1

Release Notes - February 28, 2025

Overview

This release introduces a fully functional MLX-Audio package with text-to-speech capabilities, complete with testing infrastructure and CI/CD integration via GitHub Actions.

New Features

• Text-to-Speech Generation: Added complete generation pipeline with audio output functionality
• Audio Joining: New functionality to join multiple audio segments
• Model Quantization: Added support for model quantization to improve performance
• GitHub Actions: Implemented CI/CD workflows for automated testing and deployment

Improvements

• Kokoro MLX porting: Completed refactoring of the entire model to MLX framework:
  • Text encoder with BERT implementation
  • Decoder with improved audio quality
  • Duration, indices, and alignment target prediction
  • Custom Bidirectional LSTM, Weight norm for CNNs, AdaLayerNorm and Generator layers
• SafeTensors Support: Added working implementation for SafeTensors format
• Pipeline Structure: Restructured the generation pipeline for better maintainability

Bug Fixes

• Fixed model loading mechanism
• Resolved issues with text encoder LayerNorm operation
• Fixed generator functionality
• Addressed issues in LSTM and AdaLayerNorm implementations
• Refactored and fixed ConvWeight component

Full Changelog: https://github.com/Blaizzy/mlx-audio/commits/v0.0.1