From 61e42eaf238598519675dc30e40be9630bc254f0 Mon Sep 17 00:00:00 2001 From: Kai Chappel Date: Fri, 15 May 2026 14:29:22 +1000 Subject: [PATCH] Update kokoro to v0.3.0 Co-authored-by: Kai Chappel --- kokoro/docker-compose.yml | 2 +- kokoro/umbrel-app.yml | 16 ++++++++-------- 2 files changed, 9 insertions(+), 9 deletions(-) diff --git a/kokoro/docker-compose.yml b/kokoro/docker-compose.yml index 28df1c908..3f9144eb8 100644 --- a/kokoro/docker-compose.yml +++ b/kokoro/docker-compose.yml @@ -7,6 +7,6 @@ services: APP_PORT: 8880 web: - image: ghcr.io/remsky/kokoro-fastapi-cpu:v0.2.4-master@sha256:03a5201c0327f1f1a1619ebbd9722e3b5477b0cccc588aaddb74dd89aad67f15 + image: ghcr.io/remsky/kokoro-fastapi-cpu:v0.3.0@sha256:d7df384acb57929f88a696d892c1a2e0c6b60fa934802ebec9a88626003b8e9a user: "1000:1000" restart: on-failure diff --git a/kokoro/umbrel-app.yml b/kokoro/umbrel-app.yml index 217090b0f..3a5d2f642 100644 --- a/kokoro/umbrel-app.yml +++ b/kokoro/umbrel-app.yml @@ -3,7 +3,7 @@ id: kokoro name: Kokoro tagline: An intelligent, high-quality TTS solution category: ai -version: "0.2.4-master" +version: "0.3.0" port: 8877 description: >- Kokoro is an advanced Text-to-Speech (TTS) model that delivers impressive speech quality with only 82 million parameters, making it competitive with much larger and more resource-intensive models. Despite its relatively compact architecture, Kokoro effectively transforms text into clear, natural-sounding speech, making it an excellent choice for applications relying on speech synthesis. The model has been specifically designed to ensure high efficiency and fast processing, making it suitable for both resource-constrained environments and production systems. In comparison to traditional TTS models, which often require substantial computational resources, Kokoro offers a more cost-effective and faster alternative without compromising the quality of speech output. @@ -28,15 +28,15 @@ gallery: - 3.jpg releaseNotes: >- This release includes improvements and fixes: - - Fixed custom phenomes and made them more robust - - Updated PyTorch for RTX 50 Series GPU support - - Reduced Docker image size - - Added configurable logging level with environment variable support - - Improved build stability and compatibility - - Enhanced build times through parallelization + - Improved text normalization, sentence splitting, audio pause handling, and MP3/OGG output reliability + - Added a volume multiplier setting and gpt-4o-mini-tts support + - Improved reverse proxy support for the web UI + - Fixed streaming responses with download links + - Added voice tensor safety/caching improvements and reduced per-request memory usage + - Improved Docker build stability and compatibility across CPU/GPU targets - Full release notes are available at https://github.com/remsky/Kokoro-FastAPI/releases/tag/v0.2.4-master + Full release notes are available at https://github.com/remsky/Kokoro-FastAPI/releases/tag/v0.3.0 dependencies: [] path: "/web" defaultUsername: ""