Bump to CUDA 13.0.1 + PyTorch 2.9.0, add version output on git checkouts

vllm/Dockerfile

@@ -54,8 +54,8 @@ ENV PATH=${CUDA_HOME}/cuda/bin:${PATH}
 RUN apt-get update && apt install -y wget
 
 RUN uv pip install numpy==2.0.0
-# Install pytorch nightly
-RUN uv pip install torch==2.7.0+cu128 torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu128
+# Install PyTorch 2.9.0 with CUDA 13.0
+RUN uv pip install torch==2.9.0+cu130 torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu130
 
 FROM base AS build-base
 RUN mkdir /wheels
@@ -95,7 +95,12 @@ FROM build-base AS build-lmcache
 RUN git clone https://github.com/LMCache/LMCache.git && \
     cd LMCache && \
     git checkout dev && \
-# PyTorch version is dated in LMCache
+    echo "\n\n========================================" && \
+    echo ">>> BUILDING LMCACHE FROM:" && \
+    echo ">>> BRANCH: $(git rev-parse --abbrev-ref HEAD)" && \
+    echo ">>> COMMIT: $(git rev-parse HEAD)" && \
+    echo ">>> DATE:   $(git log -1 --format=%cd --date=short)" && \
+    echo "========================================\n\n" && \
     sed -i '/torch/d' pyproject.toml && \
     uv pip install setuptools_scm && \
     MAX_JOBS=8 python -m build --wheel --no-isolation && \
@@ -209,3 +214,8 @@ RUN uv pip uninstall pynvml && uv pip install nvidia-ml-py
 # API server entrypoint
 # ENTRYPOINT ["vllm", "serve"]
 CMD ["/bin/bash"]
+p uninstall pynvml && uv pip install nvidia-ml-py
+
+# API server entrypoint
+# ENTRYPOINT ["vllm", "serve"]
+CMD ["/bin/bash"]