c86e323d7a
Quality-first multi-pass audiobook generation from English manuscripts. Includes: - umbraaudio/ core package (pipeline, config, voices, web review UI) - scripts/ CLI utilities + model download script - tts_server/ remote TTS server - specs/ design docs, tests/, pyproject.toml + uv.lock - Qwen3-TTS git submodule (pinned 022e286) with local demo.py UI patch saved at scripts/qwen3_tts_demo_umbra_patch.patch - .env with runtime config/secrets Excluded (see .gitignore): models/ (downloaded via scripts/download_models.sh), output/, logs/, .venv, .umbra runtime data, and book projects (DealHunterShip_*), which are committed separately.
27 lines
1.3 KiB
Bash
27 lines
1.3 KiB
Bash
#!/usr/bin/env bash
|
|
# Downloads the Qwen3-TTS model weights required by UmbraAudio.
|
|
#
|
|
# These models are stock HuggingFace downloads (~9 GB). They are NOT tracked
|
|
# in git; run this script once after cloning to reconstruct the models/ tree.
|
|
#
|
|
# Usage:
|
|
# bash scripts/download_models.sh [models_dir]
|
|
#
|
|
# Defaults to ./models relative to the UmbraAudio repo root.
|
|
set -euo pipefail
|
|
|
|
REPO_ROOT="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
|
|
MODELS_DIR="${1:-$REPO_ROOT/models}"
|
|
|
|
mkdir -p "$MODELS_DIR"
|
|
|
|
# Qwen3-TTS models used by the configured engines (see umbraaudio/config.py and .env).
|
|
huggingface-cli download "Qwen/Qwen3-TTS-12Hz-0.6B-Base" --local-dir "$MODELS_DIR/Qwen3-TTS-12Hz-0.6B-Base"
|
|
huggingface-cli download "Qwen/Qwen3-TTS-12Hz-0.6B-CustomVoice" --local-dir "$MODELS_DIR/Qwen3-TTS-12Hz-0.6B-CustomVoice"
|
|
huggingface-cli download "Qwen/Qwen3-TTS-12Hz-1.7B-Base" --local-dir "$MODELS_DIR/Qwen3-TTS-12Hz-1.7B-Base"
|
|
huggingface-cli download "Qwen/Qwen3-TTS-12Hz-1.7B-CustomVoice" --local-dir "$MODELS_DIR/Qwen3-TTS-12Hz-1.7B-CustomVoice"
|
|
huggingface-cli download "Qwen/Qwen3-TTS-12Hz-1.7B-VoiceDesign" --local-dir "$MODELS_DIR/Qwen3-TTS-12Hz-1.7B-VoiceDesign"
|
|
huggingface-cli download "Qwen/Qwen3-TTS-Tokenizer-12Hz" --local-dir "$MODELS_DIR/Qwen3-TTS-Tokenizer-12Hz"
|
|
|
|
echo "Models downloaded to $MODELS_DIR"
|