Attempt to fix xformers build

Signed-off-by: Huy Do <huydhn@gmail.com>

Author: huydhn

docker/Dockerfile

@@ -364,8 +364,13 @@ RUN --mount=type=bind,from=build,src=/workspace/dist,target=/vllm-workspace/dist
         --extra-index-url ${PYTORCH_CUDA_INDEX_BASE_URL}/cu$(echo $CUDA_VERSION | cut -d. -f1,2 | tr -d '.')
 
 # TODO (huydhn): Remove this once xformers is released for 2.8.0
-RUN --mount=type=cache,target=/root/.cache/uv \
+RUN --mount=type=cache,target=/root/.cache/uv bash - <<'BASH'
+    . /etc/environment
+    export TORCH_CUDA_ARCH_LIST='7.0 7.5 8.0 8.9 9.0 10.0 12.0'
     uv pip install --system --no-build-isolation "git+https://github.com/facebookresearch/xformers@v0.0.30"
+    # DEBUG
+    python -m xformers.info
+BASH
 
 # If we need to build FlashInfer wheel before its release:
 # $ # Note we remove 7.0 from the arch list compared to the list below, since FlashInfer only supports sm75+