Containerfile.build

FROM docker.io/nvidia/cuda:12.6.2-devel-ubi9

ARG OLLAMA_VERSION
ARG GOLANG_VERSION=1.22.1
ARG CMAKE_VERSION=3.22.1

ENV VERSION ${OLLAMA_VERSION}

RUN dnf -y update \
    && dnf install -y --nodocs pciutils git

# The cmake and golang dependencies require special logic
COPY ./scripts/rh_linux_deps.sh /
RUN CMAKE_VERSION=${CMAKE_VERSION} GOLANG_VERSION=${GOLANG_VERSION} sh /rh_linux_deps.sh

RUN dnf clean all && rm -rf /var/cache/* \
    && git clone --depth 1 --branch $VERSION https://github.com/ollama/ollama.git

#USER root
WORKDIR /ollama

RUN go generate ./... \
    && go build . \
    && ls -l /ollama

# RUNTIME CONTAINER
FROM docker.io/nvidia/cuda:12.6.2-runtime-ubi9
ARG OLLAMA_VERSION

RUN dnf -y update \
    && dnf install -y --nodocs pciutils \
    && dnf clean all && rm -rf /var/cache/*

COPY --from=0 /ollama/ollama /usr/local/bin/ollama
COPY VERSION /OLLAMA_VERSION

LABEL maintainer=william.caban@gmail.com \
      io.k8s.display-name="Ollama AI" \
      summary="ollama.ai - tool for running LLMs locally"

#ENV PYTHONDONTWRITEBYTECODE 1
#ENV PYTHONUNBUFFERED 1

EXPOSE 11434
ENV OLLAMA_HOST 0.0.0.0
ENV OLLAMA_VERSION ${OLLAMA_VERSION}-ubi9
#ENV OLLAMA_ORIGINS  http://localhost:*,http://0.0.0.0:*
ENV OLLAMA_MODELS /.ollama/models

ENTRYPOINT ["/usr/local/bin/ollama"]
CMD ["serve"]