Installation guide for CUDA 11 with latest version. #6087
Unanswered
for-just-we
asked this question in
Q&A
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
The installation method:
pip install "sglang[all]>=0.4.6.post2"
is for CUDA 12. When I try to install for CUDA 11 with command:pip install torch==2.6.0 torchvision==0.21.0 torchaudio==2.6.0 --index-url https://download.pytorch.org/whl/cu118 pip install cuda-bindings==11.8.6 pip install cuda-python==11.8.5 pip install cupy-cuda11x pip install flashinfer-python==0.2.5 --index-url https://flashinfer.ai/whl/cu118/torch2.6 pip install sgl-kernel==0.1.1 -i https://docs.sglang.ai/whl/cu118 pip install "sglang[all]>=0.4.6.post2"
Then run command:
python -m sglang.launch_server
. And have following error trace.It seems the
common_ops.abi3.so
in sgl_kernel-0.1.1-cp39-abi3-manylinux2014_x86_64.whl does not contain definition of function: ApplyTokenBitmaskInplace .So may I ask if:
common_ops.abi3.so
be fixed?Beta Was this translation helpful? Give feedback.
All reactions