241 lines
5.0 KiB
Markdown
241 lines
5.0 KiB
Markdown
# ViGent 手动部署指南
|
||
|
||
## 服务器信息
|
||
|
||
| 配置 | 规格 |
|
||
|------|------|
|
||
| 服务器 | Dell PowerEdge R730 |
|
||
| CPU | 2× Intel Xeon E5-2680 v4 (56 线程) |
|
||
| 内存 | 192GB DDR4 |
|
||
| GPU 0 | NVIDIA RTX 3090 24GB |
|
||
| GPU 1 | NVIDIA RTX 3090 24GB (用于 LatentSync) |
|
||
| 部署路径 | `/home/rongye/ProgramFiles/ViGent2` |
|
||
|
||
---
|
||
|
||
## 步骤 1: 环境检查
|
||
|
||
```bash
|
||
# 检查 GPU
|
||
nvidia-smi
|
||
|
||
# 检查 Python 版本 (需要 3.10+)
|
||
python3 --version
|
||
|
||
# 检查 Node.js 版本 (需要 18+)
|
||
node --version
|
||
|
||
# 检查 FFmpeg
|
||
ffmpeg -version
|
||
```
|
||
|
||
如果缺少 FFmpeg:
|
||
```bash
|
||
sudo apt update
|
||
sudo apt install ffmpeg
|
||
```
|
||
|
||
---
|
||
|
||
## 步骤 2: 创建目录结构
|
||
|
||
```bash
|
||
mkdir -p /home/rongye/ProgramFiles/ViGent2
|
||
cd /home/rongye/ProgramFiles/ViGent2
|
||
```
|
||
|
||
将项目文件复制到该目录。
|
||
|
||
---
|
||
|
||
## 步骤 3: 安装后端依赖
|
||
|
||
```bash
|
||
cd /home/rongye/ProgramFiles/ViGent2/backend
|
||
|
||
# 创建虚拟环境
|
||
python3 -m venv venv
|
||
source venv/bin/activate
|
||
|
||
# 安装 PyTorch (CUDA 12.1)
|
||
pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu121
|
||
|
||
# 安装 Python 依赖
|
||
pip install -r requirements.txt
|
||
|
||
# 安装 Playwright 浏览器（社交发布需要）
|
||
playwright install chromium
|
||
```
|
||
|
||
---
|
||
|
||
## 步骤 4: 部署 AI 模型 (LatentSync 1.6)
|
||
|
||
> ⚠️ **重要**：LatentSync 需要独立的 Conda 环境和 **~18GB VRAM**。请**不要**直接安装在后端环境中。
|
||
|
||
请参考详细的独立部署指南：
|
||
**[LatentSync 部署指南](../models/LatentSync/DEPLOY.md)**
|
||
|
||
该指南包含以下关键步骤，请务必严格按照文档操作：
|
||
1. 创建独立的 `latentsync` Conda 环境
|
||
2. 安装 PyTorch 2.5.1 和相关依赖
|
||
3. 下载模型权重 (HuggingFace CLI)
|
||
4. 复制核心推理代码
|
||
5. 验证推理脚本
|
||
|
||
确保 LatentSync 部署成功后，再继续后续步骤。
|
||
|
||
---
|
||
|
||
## 步骤 5: 启动 LatentSync 常驻加速服务 (可选)
|
||
|
||
为了消除每次生成视频时的 30-40秒 模型加载时间，建议启动常驻服务：
|
||
|
||
```bash
|
||
cd /home/rongye/ProgramFiles/ViGent2/models/LatentSync
|
||
|
||
# 后台启动服务 (自动读取 backend/.env 中的 GPU 配置)
|
||
nohup python -m scripts.server > server.log 2>&1 &
|
||
```
|
||
|
||
---
|
||
|
||
## 步骤 7: 配置环境变量
|
||
|
||
```bash
|
||
cd /home/rongye/ProgramFiles/ViGent2/backend
|
||
|
||
# 复制配置模板 (默认配置已经就绪)
|
||
cp .env.example .env
|
||
```
|
||
|
||
> 💡 **说明**：`.env.example` 已包含正确的 LatentSync 默认配置，直接复制即可使用。
|
||
> 如需自定义，可编辑 `.env` 修改以下参数：
|
||
|
||
| 配置项 | 默认值 | 说明 |
|
||
|--------|--------|------|
|
||
| `LATENTSYNC_GPU_ID` | 1 | GPU 选择 (0 或 1) |
|
||
| `LATENTSYNC_USE_SERVER` | false | 设为 true 以启用常驻服务加速 |
|
||
| `LATENTSYNC_INFERENCE_STEPS` | 20 | 推理步数 (20-50) |
|
||
| `LATENTSYNC_GUIDANCE_SCALE` | 1.5 | 引导系数 (1.0-3.0) |
|
||
| `DEBUG` | true | 生产环境改为 false |
|
||
|
||
---
|
||
|
||
## 步骤 8: 安装前端依赖
|
||
|
||
```bash
|
||
cd /home/rongye/ProgramFiles/ViGent2/frontend
|
||
|
||
# 安装依赖
|
||
npm install
|
||
```
|
||
|
||
---
|
||
|
||
## 步骤 9: 测试运行
|
||
|
||
### 启动后端
|
||
|
||
```bash
|
||
cd /home/rongye/ProgramFiles/ViGent2/backend
|
||
source venv/bin/activate
|
||
uvicorn app.main:app --host 0.0.0.0 --port 8006
|
||
```
|
||
|
||
### 启动前端 (新开终端)
|
||
|
||
```bash
|
||
cd /home/rongye/ProgramFiles/ViGent2/frontend
|
||
npm run dev -- -H 0.0.0.0 --port 3002
|
||
```
|
||
|
||
---
|
||
|
||
## 步骤 10: 验证
|
||
|
||
1. 访问 http://服务器IP:3002 查看前端
|
||
2. 访问 http://服务器IP:8006/docs 查看 API 文档
|
||
3. 上传测试视频，生成口播视频
|
||
|
||
---
|
||
|
||
## 使用 systemd 管理服务 (可选)
|
||
|
||
### 后端服务
|
||
|
||
创建 `/etc/systemd/system/vigent2-backend.service`:
|
||
```ini
|
||
[Unit]
|
||
Description=ViGent2 Backend API
|
||
After=network.target
|
||
|
||
[Service]
|
||
Type=simple
|
||
User=rongye
|
||
WorkingDirectory=/home/rongye/ProgramFiles/ViGent2/backend
|
||
Environment="PATH=/home/rongye/ProgramFiles/ViGent2/backend/venv/bin"
|
||
ExecStart=/home/rongye/ProgramFiles/ViGent2/backend/venv/bin/uvicorn app.main:app --host 0.0.0.0 --port 8006
|
||
Restart=always
|
||
|
||
[Install]
|
||
WantedBy=multi-user.target
|
||
```
|
||
|
||
### 前端服务
|
||
|
||
创建 `/etc/systemd/system/vigent2-frontend.service`:
|
||
```ini
|
||
[Unit]
|
||
Description=ViGent2 Frontend
|
||
After=network.target
|
||
|
||
[Service]
|
||
Type=simple
|
||
User=rongye
|
||
WorkingDirectory=/home/rongye/ProgramFiles/ViGent2/frontend
|
||
ExecStart=/usr/bin/npm run start
|
||
Restart=always
|
||
|
||
[Install]
|
||
WantedBy=multi-user.target
|
||
```
|
||
|
||
### 启用服务
|
||
|
||
```bash
|
||
sudo systemctl daemon-reload
|
||
sudo systemctl enable vigent2-backend vigent2-frontend
|
||
sudo systemctl start vigent2-backend vigent2-frontend
|
||
```
|
||
|
||
---
|
||
|
||
## 故障排除
|
||
|
||
### GPU 不可用
|
||
|
||
```bash
|
||
# 检查 CUDA
|
||
nvidia-smi
|
||
python3 -c "import torch; print(torch.cuda.is_available())"
|
||
```
|
||
|
||
### 端口被占用
|
||
|
||
```bash
|
||
# 查看端口占用
|
||
sudo lsof -i :8006
|
||
sudo lsof -i :3002
|
||
```
|
||
|
||
### 查看日志
|
||
|
||
```bash
|
||
# 后端日志
|
||
journalctl -u vigent2-backend -f
|
||
|
||
# 前端日志
|
||
journalctl -u vigent2-frontend -f
|
||
```
|