Move docker stuff to a dedicated folder

docker/Dockerfile.llamacpp

@@ -0,0 +1,23 @@
+FROM ghcr.io/ggml-org/llama.cpp:server-cuda
+
+# Install curl for downloading llamactl
+RUN apt-get update && apt-get install -y curl && rm -rf /var/lib/apt/lists/*
+
+# Download and install the latest llamactl release
+RUN LATEST_VERSION=$(curl -s https://api.github.com/repos/lordmathis/llamactl/releases/latest | grep '"tag_name":' | sed -E 's/.*"([^"]+)".*/\1/') && \
+    curl -L "https://github.com/lordmathis/llamactl/releases/download/${LATEST_VERSION}/llamactl-${LATEST_VERSION}-linux-amd64.tar.gz" | tar -xz && \
+    mv llamactl /usr/local/bin/ && \
+    chmod +x /usr/local/bin/llamactl
+
+# Set working directory
+RUN mkdir -p /data
+WORKDIR /data
+
+# Expose the default llamactl port
+EXPOSE 8080
+
+ENV LLAMACTL_LLAMACPP_COMMAND=/app/llama-server
+ENV LD_LIBRARY_PATH="/app:/usr/local/lib:/usr/lib"
+
+# Set llamactl as the entrypoint
+ENTRYPOINT ["llamactl"]

docker/Dockerfile.vllm

@@ -0,0 +1,20 @@
+FROM vllm/vllm-openai:latest
+
+# Install curl for downloading llamactl
+RUN apt-get update && apt-get install -y curl && rm -rf /var/lib/apt/lists/*
+
+# Download and install the latest llamactl release
+RUN LATEST_VERSION=$(curl -s https://api.github.com/repos/lordmathis/llamactl/releases/latest | grep '"tag_name":' | sed -E 's/.*"([^"]+)".*/\1/') && \
+    curl -L "https://github.com/lordmathis/llamactl/releases/download/${LATEST_VERSION}/llamactl-${LATEST_VERSION}-linux-amd64.tar.gz" | tar -xz && \
+    mv llamactl /usr/local/bin/ && \
+    chmod +x /usr/local/bin/llamactl
+
+# Set working directory
+RUN mkdir -p /data
+WORKDIR /data
+
+# Expose the default llamactl port
+EXPOSE 8080
+
+# Set llamactl as the entrypoint
+ENTRYPOINT ["llamactl"]

docker/docker-compose.yml

@@ -0,0 +1,56 @@
+version: '3.8'
+
+services:
+  llamactl-llamacpp:
+    build:
+      context: ..
+      dockerfile: docker/Dockerfile.llamacpp
+    image: llamactl:llamacpp-cuda
+    container_name: llamactl-llamacpp
+    ports:
+      - "8080:8080"
+    volumes:
+      - ./data/llamacpp:/data
+      - ./models:/models  # Mount models directory
+      - ~/.cache/llama.cpp:/root/.cache/llama.cpp  # Llama.cpp cache
+    environment:
+      # Set data directory for persistence
+      - LLAMACTL_DATA_DIR=/data
+      # Enable Docker mode for nested containers (if needed)
+      - LLAMACTL_LLAMACPP_DOCKER_ENABLED=false
+    deploy:
+      resources:
+        reservations:
+          devices:
+            - driver: nvidia
+              count: all
+              capabilities: [gpu]
+    restart: unless-stopped
+
+  llamactl-vllm:
+    build:
+      context: ..
+      dockerfile: docker/Dockerfile.vllm
+    image: llamactl:vllm-cuda
+    container_name: llamactl-vllm
+    ports:
+      - "8081:8080"  # Use different port to avoid conflicts
+    volumes:
+      - ./data/vllm:/data
+      - ./models:/models  # Mount models directory
+      - ~/.cache/huggingface:/root/.cache/huggingface  # HuggingFace cache
+    environment:
+      # Set data directory for persistence
+      - LLAMACTL_DATA_DIR=/data
+      # Enable Docker mode for nested containers (if needed)
+      - LLAMACTL_VLLM_DOCKER_ENABLED=false
+      # vLLM specific environment variables
+      - CUDA_VISIBLE_DEVICES=all
+    deploy:
+      resources:
+        reservations:
+          devices:
+            - driver: nvidia
+              count: all
+              capabilities: [gpu]
+    restart: unless-stopped