fix: add missing argument to CMD in Dockerfile

Dockerfile

@@ -11,9 +11,9 @@ RUN apt-get update && apt-get install -y \
 
 WORKDIR /build
 
-# Pin a known-good release that supports --jinja (added after b3600).
-# Bump this periodically; check https://github.com/ggml-org/llama.cpp/releases
-RUN git clone --depth 1 --branch b4996 https://github.com/ggml-org/llama.cpp.git
+# Pin a known-good release that supports --jinja.
+# Verify current tags at https://github.com/ggml-org/llama.cpp/releases before bumping.
+RUN git clone --depth 1 --branch b10107 https://github.com/ggml-org/llama.cpp.git
 
 WORKDIR /build/llama.cpp
 
@@ -40,9 +40,6 @@ WORKDIR /app
 
 COPY --from=builder /build/llama.cpp/build/bin/llama-server /app/llama-server
 
-# Qwen2.5-3B is a better fit than 7B for a 2-thread CPU-only box —
-# noticeably faster time-to-first-token for tool-calling latency.
-# Override with --build-arg MODEL_URL=... if you want to go back to 7B.
 ARG MODEL_URL=https://huggingface.co/bartowski/Qwen2.5-3B-Instruct-GGUF/resolve/main/Qwen2.5-3B-Instruct-Q4_K_M.gguf
 RUN curl -L -o /app/model.gguf "${MODEL_URL}"