diff --git a/llama.cpp-base/Dockerfile b/llama.cpp-base/Dockerfile index 87edab2..b0ea175 100644 --- a/llama.cpp-base/Dockerfile +++ b/llama.cpp-base/Dockerfile @@ -34,7 +34,7 @@ ENV LLAMA_CURL=1 ENV LLAMA_CUDA_FA_ALL_QUANTS=1 ENV GGML_CUDA_FA_ALL_QUANTS=1 -RUN make -j 12 +RUN make -j 12 llama-batched-bench llama-server FROM ${CUDA_BASE_CONTAINER} as runtime