Alibaba Cloud Mirror Quick Start#

This guide is intended for users in mainland China, using Alibaba Cloud (Aliyun) mirrors to accelerate the FluxVLA installation.

1) Create a Conda Environment#

conda create -n fluxvla python=3.10 -y
conda activate fluxvla

2) Configure pip to Use Alibaba Cloud Mirror#

pip config set global.index-url https://mirrors.aliyun.com/pypi/simple/
pip config set install.trusted-host mirrors.aliyun.com

3) Install PyTorch (Official CUDA Wheel Source)#

pip install torch==2.6.0 torchvision==0.21.0 torchaudio==2.6.0 --index-url https://download.pytorch.org/whl/cu124

4) Install flash-attention#

pip install flash-attn==2.5.5 --no-build-isolation --find-links https://github.com/Dao-AILab/flash-attention/releases

If your network or build environment is restricted, refer to the source installation method in :doc:zero.

5) Install av#

conda install -c conda-forge av=14.4.0

6) Install FluxVLA#

pip install -r requirements.txt
pip install -e . --no-build-isolation

7) Fix numpy Version (Required)#

pip uninstall numpy
pip install numpy==1.26.4

Optional: Configure the Online Evaluation Environment#

To evaluate LIBERO on devices without ray-tracing capability (e.g., A100), refer to the EGL Device GPU Rendering Configuration.

export MUJOCO_GL=egl
sudo apt install libegl-dev libgl1-mesa-dev libx11-dev libglew-dev libosmesa6-dev