STT (Silero VAD + Whisper via sherpa-onnx), Chatterbox TTS HTTP server, query completeness classifier (Ollama), multi-voice demo scripts, and planning docs. Kept as reference; clean rewrite planned in separate repos. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
64 lines
1.8 KiB
Bash
Executable File
64 lines
1.8 KiB
Bash
Executable File
#!/usr/bin/env bash
|
|
# Download ONNX models for the voice experiment.
|
|
#
|
|
# Usage:
|
|
# bash download-models.sh # default: whisper base.en
|
|
# WHISPER_MODEL=large-v3-turbo bash download-models.sh
|
|
#
|
|
# After downloading, run: node demo.mjs
|
|
|
|
set -euo pipefail
|
|
|
|
WHISPER_MODEL="${WHISPER_MODEL:-base.en}"
|
|
SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
|
|
MODELS_DIR="${SCRIPT_DIR}/models"
|
|
|
|
mkdir -p "${MODELS_DIR}"
|
|
cd "${MODELS_DIR}"
|
|
|
|
echo "==> models dir: ${MODELS_DIR}"
|
|
|
|
# --- Silero VAD (~2 MB) ---
|
|
if [ ! -f silero_vad.onnx ]; then
|
|
echo "==> downloading silero_vad.onnx"
|
|
curl -L --progress-bar -o silero_vad.onnx \
|
|
'https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/silero_vad.onnx'
|
|
else
|
|
echo "==> silero_vad.onnx already present"
|
|
fi
|
|
|
|
# --- Whisper STT ---
|
|
WHISPER_DIR="sherpa-onnx-whisper-${WHISPER_MODEL}"
|
|
if [ ! -d "${WHISPER_DIR}" ]; then
|
|
echo "==> downloading whisper-${WHISPER_MODEL}"
|
|
curl -L --progress-bar -o "${WHISPER_DIR}.tar.bz2" \
|
|
"https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/sherpa-onnx-whisper-${WHISPER_MODEL}.tar.bz2"
|
|
echo "==> extracting..."
|
|
tar xf "${WHISPER_DIR}.tar.bz2"
|
|
else
|
|
echo "==> whisper-${WHISPER_MODEL} already present"
|
|
fi
|
|
|
|
# --- Kokoro TTS (~310 MB) ---
|
|
if [ ! -d kokoro-en-v0_19 ]; then
|
|
echo "==> downloading kokoro-en-v0_19 (English TTS)"
|
|
curl -L --progress-bar -o kokoro-en-v0_19.tar.bz2 \
|
|
'https://github.com/k2-fsa/sherpa-onnx/releases/download/tts-models/kokoro-en-v0_19.tar.bz2'
|
|
echo "==> extracting..."
|
|
tar xf kokoro-en-v0_19.tar.bz2
|
|
else
|
|
echo "==> kokoro-en-v0_19 already present"
|
|
fi
|
|
|
|
echo ""
|
|
echo "==> all models ready"
|
|
echo ""
|
|
echo "Next steps:"
|
|
echo " cd ${SCRIPT_DIR}"
|
|
echo " npm install"
|
|
echo " node demo.mjs"
|
|
echo ""
|
|
echo "To use a better STT model:"
|
|
echo " WHISPER_MODEL=large-v3-turbo bash download-models.sh"
|
|
echo " WHISPER_MODEL=large-v3-turbo node demo.mjs"
|