fix: add ldflags for version, remove privileged, enable flash attention

ai/compose.yml

@@ -63,7 +63,6 @@ services:
       dockerfile: Dockerfile
     image: ollama/ollama:rocm-gfx906
     container_name: ollama
-    privileged: true
     tty: true
     restart: always 
     ports:
@@ -81,7 +80,7 @@ services:
       - HSA_ENABLE_SDMA=0 
       - OLLAMA_HOST=0.0.0.0
       - OLLAMA_DEBUG=1
-      - OLLAMA_FLASH_ATTENTION=0
+      - OLLAMA_FLASH_ATTENTION=1
       - OLLAMA_NUM_PARALLEL=2
     devices:
       # Map the render nodes and KFD for ROCm to work inside the container

ai/ollama/Dockerfile

@@ -74,7 +74,7 @@ RUN mkdir -p build-hip && \
 
 # Step 3: Build Go binary (GCC for CGo linking)
 ENV CGO_ENABLED=1
-RUN go build -trimpath -o /build/dist/ollama .
+RUN go build -trimpath -ldflags="-X=github.com/ollama/ollama/version.Version=${OLLAMA_VERSION}" -o /build/dist/ollama .
 
 # ---------- Runtime image ----------
 FROM ubuntu:24.04