Установака TTS модели локально

## Руководство по ручной установке для VibeVoice WebUI

Step 1: Prerequisites

Before you begin, make sure you have installed:

1. Git – https://git-scm.com/download/win
2. Python 3.10 – https://www.python.org/downloads/release/python-31011/
3. NVIDIA CUDA Toolkit 12.8 (or newer) – https://developer.nvidia.com/cuda-downloads

Step 2: Download and Install VibeVoiceTTS

- Open cmd or PowerShell in the folder where you want to install.
- Clone the repo:

git clone https://github.com/SUP3RMASS1VE/VibeVoiceTTS.git
cd VibeVoiceTTS

Step 3: Create Python Virtual Environment

Внутри папки VibeVoiceTTS:

python -m venv .venv
.venv\Scripts\activate

Step 4: Install Dependencies

Выполните следующие команды:

pip install uv
uv pip install torch==2.7.0 torchvision torchaudio --index-url https://download.pytorch.org/whl/cu128
uv pip install triton-windows
uv pip install https://github.com/petermg/flash_attn_windows/releases/download/01/flash_attn-2.7.4.post1+cu128.torch270-cp310-cp310-win_amd64.whl
uv pip install -e .

Это установит Torch с поддержкой CUDA 12.8, FlashAttention и все зависимости VibeVoice.

Step 5: Download the Launcher

Загрузите скрипт запуска помощника в папку VibeVoice:

curl -L -o LAUNCHER_VibeVoice.bat "https://huggingface.co/Aitrepreneur/FLX/resolve/main/LAUNCHER_VibeVoice.bat?download=true"

Step 6: Run the WebUI

Запустите WebUI с помощью лаунчера

------------------------------------------------------------------------------------------------------------------------------------

## Руководство по ручной установке для RVC WebUI (Windows)

Step 1: Prerequisites

Make sure you have installed:

- 7-Zip – https://www.7-zip.org/
- Git – https://git-scm.com/download/win

Step 2: Download and Extract RVC WebUI

Определите тип вашего графического процессора:

1. NVIDIA GPU:
https://huggingface.co/lj1995/VoiceConversionWebUI/resolve/main/RVC1006Nvidia.7z
2. AMD / Intel GPU:
https://huggingface.co/lj1995/VoiceConversionWebUI/resolve/main/RVC1006AMD_Intel.7z

Download and extract with 7-Zip. Ваша папка должна выглядеть так:

RVC1006Nvidia/
├── go-web.bat
├── requirements.txt
├── infer-web.py
└── ...

Step 3: Run the WebUI

Double-click go-web.bat

-----------------------------------------------------------------------------------------------------------------------

## Руководство по ручной установке VibeVoice WebUI (RunPod)

Step 1: Prerequisites

1. Create a RunPod account: https://bit.ly/runpodAi
2. Deploy a pod with the PyTorch 2.1
3. Open a terminal inside /workspace

Step 2: Install System Packages

apt-get update
apt-get install -y build-essential ninja-build cmake ffmpeg git

Step 3: Clone and Enter Repo

git clone https://github.com/SUP3RMASS1VE/VibeVoiceTTS.git
cd VibeVoiceTTS

Step 4: Create Python Environment

python3 -m venv .venv
source .venv/bin/activate
pip install -U pip setuptools wheel packaging

Step 5: Install Dependencies

pip install uv
uv pip install torch==2.8.0 torchvision torchaudio --index-url https://download.pytorch.org/whl/cu128
uv pip install "flash-attn==2.8.3" --no-build-isolation
uv pip install -e .

Step 6: Run the WebUI

python demo/gradio_demo.py --share


Open the provided public link in your browser.

--------------------------------------------------------------------------------------------------------------------------


## Руководство по ручной установке для RVC WebUI (RunPod)

Step 1: Prerequisites

1. Create a RunPod account: https://bit.ly/runpodAi
2. Deploy a pod with the PyTorch 2.1
3. Open a terminal inside /workspace

Step 2: Update System

apt update && apt upgrade -y

Step 3: Download and Extract RVC

wget https://huggingface.co/lj1995/VoiceConversionWebUI/resolve/main/RVC1006Nvidia.7z
apt install p7zip-full -y
7z x RVC1006Nvidia.7z
cd RVC1006Nvidia

Step 4: Install Dependencies

apt update && apt upgrade -y
pip install --force-reinstall -v "av==11.0.0"
pip install matplotlib==3.7.0
pip install -r requirements.txt

Step 5: Run the WebUI

python infer-web.py

----------------------------------------------------------------------------------

## Руководство по ручной установке для ComfyUI (Windows)

Step 1: Prerequisites

Install:
- 7-Zip – https://www.7-zip.org/
- Git – https://git-scm.com/download/win

Step 2: Download and Extract ComfyUI

Download latest portable NVIDIA build:
https://github.com/comfyanonymous/ComfyUI/releases/latest

Extract with 7-Zip. Example folder layout:

ComfyUI_windows_portable/
├── ComfyUI/
│   ├── models/
│   └── custom_nodes/
├── python_embeded/
└── run_nvidia_gpu.bat

Step 3: Install Custom Nodes

Go to ComfyUI_windows_portable\ComfyUI\custom_nodes and run:

git clone https://github.com/ltdrdata/ComfyUI-Manager.git
git clone https://github.com/rgthree/rgthree-comfy
git clone https://github.com/Enemyx-net/VibeVoice-ComfyUI
git clone https://github.com/diodiogod/TTS-Audio-Suite


Then install dependencies for each node that has a requirements.txt:

..\..\python_embeded\python.exe -m pip install -r <node>\requirements.txt

Step 4: Run ComfyUI

Go back to ComfyUI_windows_portable and double-click:

run_nvidia_gpu.bat

---------------------------------------------------------------------------------------------

## Руководство по ручной установке для r ComfyUI (RunPod)


Step 1: Prerequisites

1. Create account at https://bit.ly/runpodAi
2. Pick a 24 GB VRAM GPU with the **aitrepreneur/comfyui** template
3. Access port **8888**
4. Watch logs:
   `tail -f workspace/logs/comfyui.log`

Step 2: Install System Packages

apt-get update
DEBIAN_FRONTEND=noninteractive apt-get install -y \
  git git-lfs curl ffmpeg portaudio19-dev libasound2-dev \
  build-essential ninja-build cmake
git lfs install

Step 3: Setup Python Virtual Environment

cd /workspace/ComfyUI
python3 -m venv venv
source venv/bin/activate

pip install -U pip uv wheel setuptools
export UV_LINK_MODE=copy
export PYTHONNOUSERSITE=1
unset PYTHONPATH

Step 4: Install Torch (CUDA 12.8 build)

uv pip install --upgrade torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu128
uv pip uninstall -y xformers

Step 5: Clone Required Custom Nodes

cd custom_nodes
git clone https://github.com/diodiogod/TTS-Audio-Suite
git clone https://github.com/Enemyx-net/VibeVoice-ComfyUI
git clone https://github.com/rgthree/rgthree-comfy

Step 6: Install Python Dependencies

Run this with uv to pin correct versions:

uv pip install \
  "numpy==2.2.6" \
  "librosa==0.11.0" \
  "soundfile>=0.12.0" \
  "sounddevice>=0.4.0" \
  "accelerate>=1.6.0" \
  "transformers==4.51.3" \
  "diffusers==0.35.1" \
  "scipy" \
  "ml-collections" \
  "peft>=0.17.0" \
  "huggingface_hub>=0.25.1" \
  "absl-py" \
  "aiortc==1.13.0" \
  "av==14.4.0" \
  "bitsandbytes==0.47.0" \
  "conformer==0.3.2" \
  "x-transformers==2.7.6" \
  "torchdiffeq==0.2.5" \
  "wandb==0.21.4" \
  "ema-pytorch==0.7.7" \
  "vocos==0.1.0" \
  "monotonic-alignment-search==0.2.0" \
  "faiss-cpu>=1.7.4" \
  "praat-parselmouth>=0.4.6" \
  "pyworld==0.3.5" \
  "torchfcpe==0.0.4" \
  "opencv-python-headless==4.12.0.88" \
  "pillow" \
  "datasets==4.0.0" \
  "requests" \
  "dacite==1.9.2" \
  "unidecode==1.4.0" \
  "jieba==0.42.1" \
  "pypinyin==0.55.0" \
  "torchsde==0.2.6"

Step 7: Download HuBERT Model

mkdir -p custom_nodes/TTS-Audio-Suite/models/hubert
curl -L -o custom_nodes/TTS-Audio-Suite/models/hubert/hubert_base.pt \
  https://huggingface.co/lj1995/VoiceConversionWebUI/resolve/main/hubert_base.pt

mkdir -p custom_nodes/VibeVoice-ComfyUI/models
ln -sf ../TTS-Audio-Suite/models/hubert/hubert_base.pt \
  custom_nodes/VibeVoice-ComfyUI/models/hubert_base.pt

Step 8: Run ComfyUI