diff --git a/Dockerfile.llamacpp b/Dockerfile.llamacpp index ded5007..f06155b 100644 --- a/Dockerfile.llamacpp +++ b/Dockerfile.llamacpp @@ -9,16 +9,15 @@ RUN LATEST_VERSION=$(curl -s https://api.github.com/repos/lordmathis/llamactl/re mv llamactl /usr/local/bin/ && \ chmod +x /usr/local/bin/llamactl -# Create data directory for llamactl -RUN mkdir -p /data - # 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"] \ No newline at end of file diff --git a/docker-compose.yml b/docker-compose.yml index 2190c9e..7e00d58 100644 --- a/docker-compose.yml +++ b/docker-compose.yml @@ -13,8 +13,6 @@ services: - ./data/llamacpp:/data - ./models:/models # Mount models directory environment: - # Configure llamactl to use llama-server from the base image - - LLAMACTL_LLAMACPP_COMMAND=llama-server # Enable Docker mode for nested containers (if needed) - LLAMACTL_LLAMACPP_DOCKER_ENABLED=false deploy: @@ -39,9 +37,6 @@ services: - ./models:/models # Mount models directory - ~/.cache/huggingface:/root/.cache/huggingface # HuggingFace cache environment: - # Configure llamactl to use vllm from the base image - - LLAMACTL_VLLM_COMMAND=vllm - - LLAMACTL_VLLM_ARGS=serve # Enable Docker mode for nested containers (if needed) - LLAMACTL_VLLM_DOCKER_ENABLED=false # vLLM specific environment variables