Bleeding edge build: LMCache main, vLLM main, latest transformers

vllm/Dockerfile

@@ -126,7 +126,8 @@ RUN apt-get update && apt-get install -y build-essential cmake gcc && \
 # the PyPI download section.
 # ==============================================================================
 FROM build-base AS build-vllm
-ARG VLLM_REF=v0.18.1
+# Bleeding edge: build from main branch
+ARG VLLM_REF=main
 # Install ccache for faster compilation
 RUN apt-get update && apt-get install -y ccache
 RUN git clone https://github.com/vllm-project/vllm.git
@@ -185,7 +186,8 @@ RUN rm -r wheels
 RUN uv pip install pynvml pandas
 
 # Add additional packages for vLLM OpenAI
-RUN uv pip install accelerate hf_transfer modelscope bitsandbytes timm boto3 runai-model-streamer runai-model-streamer[s3] tensorizer
+# Bleeding edge: latest transformers
+RUN uv pip install accelerate hf_transfer modelscope bitsandbytes timm boto3 runai-model-streamer runai-model-streamer[s3] tensorizer transformers --upgrade
 
 # Clean uv cache
 RUN uv clean