[build] downgrade flashinfer to 0.4.1 (#1598)

Signed-off-by: AlpinDale <alpindale@gmail.com>

Author: AlpinDale

docker/Dockerfile

@@ -315,8 +315,8 @@ RUN --mount=type=bind,from=build,src=/workspace/dist,target=/aphrodite-workspace
 # Install FlashInfer pre-compiled kernel cache and binaries
 # https://docs.flashinfer.ai/installation.html
 RUN --mount=type=cache,target=/root/.cache/uv \
-    uv pip install --system flashinfer-cubin==0.5.0 \
-    && uv pip install --system flashinfer-jit-cache==0.5.0 \
+    uv pip install --system flashinfer-cubin==0.4.1 \
+    && uv pip install --system flashinfer-jit-cache==0.4.1 \
         --extra-index-url https://flashinfer.ai/whl/cu$(echo $CUDA_VERSION | cut -d. -f1,2 | tr -d '.') \
     && flashinfer show-config
 

requirements/cuda.txt

@@ -12,4 +12,4 @@ torchvision==0.24.0 # Required for phi3v processor. See https://github.com/pytor
 # https://github.com/facebookresearch/xformers/releases/tag/v0.0.31
 # xformers==0.0.32.post1; platform_system == 'Linux' and platform_machine == 'x86_64'  # Requires PyTorch >= 2.8
 # FlashInfer should be updated together with the Dockerfile
-flashinfer-python==0.5.0
+flashinfer-python==0.4.1