2025-10-22 15:26:42 +08:00
# Offline Example of vLLM-omni for Qwen2.5-omni
2025-10-30 14:37:07 +08:00
## 🛠️ Installation
2025-10-22 15:26:42 +08:00
2025-11-04 22:06:41 +08:00
### Environment setup
2025-10-30 17:16:25 +08:00
2025-10-30 14:37:07 +08:00
Use Docker to keep consistent basic environment (Optional, Recommended)
```bash
2025-11-05 10:47:30 +08:00
docker run --gpus all --ipc= host --network= host -v $source_dir :$container_dir --name $container_name -it nvcr.io/nvidia/pytorch:25.01-py3 bash
2025-10-30 14:37:07 +08:00
```
Set up basic uv environment
2025-10-22 15:26:42 +08:00
```bash
2025-10-30 17:18:19 +08:00
pip install uv
2025-10-22 15:26:42 +08:00
uv venv --python 3.12 --seed
source .venv/bin/activate
```
2025-11-04 22:06:41 +08:00
### Installation of vLLM (for users)
Now we build it based on vLLM v0.11.0. Please install it with command below.
```bash
uv pip install vllm == 0.11.0 --torch-backend= auto
```
### Installation of vLLM (for developers)
2025-11-05 10:47:30 +08:00
If you want to check or debug with source code of vLLM, install stable release version of vllm with 0.11.0 from source with pre-built wheel file.
2025-10-22 15:26:42 +08:00
2025-10-30 14:37:07 +08:00
```bash
git clone https://github.com/vllm-project/vllm.git
cd vllm
2025-11-04 22:06:41 +08:00
git checkout v0.11.0
2025-10-30 14:37:07 +08:00
```
Set up environment variables to get pre-built wheels. If there are internet problems, just download the whl file manually. And set VLLM_PRECOMPILED_WHEEL_LOCATION as your local absolute path of whl file.
2025-10-22 15:26:42 +08:00
```bash
2025-11-04 22:06:41 +08:00
export VLLM_PRECOMPILED_WHEEL_LOCATION = https://github.com/vllm-project/vllm/releases/download/v0.11.0/vllm-0.11.0-cp38-abi3-manylinux1_x86_64.whl
2025-10-30 14:37:07 +08:00
```
2025-11-04 22:06:41 +08:00
Install vllm with command below (If you have no existing PyTorch).
2025-10-30 14:37:07 +08:00
```bash
uv pip install --editable .
2025-10-22 15:26:42 +08:00
```
2025-11-04 22:06:41 +08:00
Install vllm with command below (If you already have PyTorch).
```bash
python use_existing_torch.py
uv pip install -r requirements/build.txt
uv pip install --no-build-isolation --editable .
```
2025-11-05 10:47:30 +08:00
### Verification for successful installation of vLLM
Just run the command below. If no error, it demonstrates that the installation is successfull.
```bash
python -c "import vllm._C"
```
2025-10-30 17:16:25 +08:00
### Installation of vLLM-omni
Install additional requirements for vllm-omni
```bash
git clone https://github.com/vllm-project/vllm-omni.git
cd vllm_omni
uv pip install -r requirements.txt
```
## Run examples (Qwen2.5-omni)
2025-10-22 15:26:42 +08:00
Get into the example folder
```bash
2025-10-30 17:16:25 +08:00
cd examples/offline_inference/qwen_2_5_omni
2025-10-22 15:26:42 +08:00
```
Modify PYTHONPATH in run.sh as your path of vllm_omni. Then run.
```bash
bash run.sh
```
The output audio is saved in ./output_audio