Ryan Malloy bf0dfa7a5e Decode SNAC in single pass to fix audio artifacts ...

SNAC's convolutional decoder has a receptive field spanning multiple
frames. Batched decode (28 tokens at a time) created boundary
discontinuities that produced muddled audio. Since pw-play waits for
the full WAV anyway, single-pass decode costs only ~2s extra and
produces clean audio.

2026-02-21 21:51:42 -07:00

src/tts_mcp

Decode SNAC in single pass to fix audio artifacts

2026-02-21 21:51:42 -07:00

.gitignore

Initial TTS MCP server with 3 engines

2026-02-20 18:10:15 -07:00

docker-compose.yml

Replace Ollama with llama-server for 15x Orpheus throughput

2026-02-21 21:33:23 -07:00

Dockerfile

Add Docker deployment with Caddy HTTPS reverse proxy

2026-02-20 20:47:22 -07:00

llama-server.Dockerfile

Replace Ollama with llama-server for 15x Orpheus throughput

2026-02-21 21:33:23 -07:00

Makefile

Replace Ollama with llama-server for 15x Orpheus throughput

2026-02-21 21:33:23 -07:00

orpheus_tts.py

Initial TTS MCP server with 3 engines

2026-02-20 18:10:15 -07:00

pyproject.toml

Replace Ollama with llama-server for 15x Orpheus throughput

2026-02-21 21:33:23 -07:00

uv.lock

Replace Ollama with llama-server for 15x Orpheus throughput

2026-02-21 21:33:23 -07:00

Description

No description provided

946 KiB

Languages

Python 95.9%

Dockerfile 2%

Swift 1.4%

Makefile 0.7%