Closed
Description
Your current environment
My python environment is 3.11.
How you are installing vllm
I have a docker image with the following commands:
RUN cd /opt/vllm && curl -sLO "https://github.com/vllm-project/vllm/archive/refs/tags/v${VLLM_VERSION}.zip" && unzip v${VLLM_VERSION}.zip
WORKDIR /opt/vllm/vllm-${VLLM_VERSION}
RUN pip uninstall torch torch-xla -y
RUN sudo apt-get install libopenblas-base libopenmpi-dev libomp-dev -y
RUN pip install -r requirements-tpu.txt
RUN VLLM_TARGET_DEVICE="tpu" python3 setup.py develop
I get the following error:
=> ERROR [24/38] RUN pip install -r requirements-tpu.txt 1.2s
------
> [24/38] RUN pip install -r requirements-tpu.txt:
#0 0.902 Looking in indexes: https://pypi.org/simple, https://download.pytorch.org/whl/nightly/cpu
#0 0.902 Looking in links: https://storage.googleapis.com/libtpu-releases/index.html, https://storage.googleapis.com/jax-releases/jax_nightly_releases.html, https://storage.googleapis.com/jax-releases/jaxlib_nightly_releases.html
#0 0.903 Ignoring fastapi: markers 'python_version < "3.9"' don't match your environment
#0 0.904 Ignoring six: markers 'python_version > "3.11"' don't match your environment
#0 0.904 Ignoring setuptools: markers 'python_version > "3.11"' don't match your environment
#0 0.910 ERROR: torch_xla-2.6.0.dev20241126-cp310-cp310-linux_x86_64.whl is not a supported wheel on this platform.
I assume this is because the torch xla version is for python 3.10. I wonder if there is a solution to this by changing the requirements tpu txt to not hardcode the 3.10 version?
Before submitting a new issue...
- Make sure you already searched for relevant issues, and asked the chatbot living at the bottom right corner of the documentation page, which can answer lots of frequently asked questions.