Skip to content

Commit

Permalink
update dockerfile merlin base container (#781)
Browse files Browse the repository at this point in the history
  • Loading branch information
@jperez999
jperez999 authored Dec 29, 2022
1 parent 0bb715f commit efa8215
Showing 1 changed file with 35 additions and 69 deletions.
104 changes: 35 additions & 69 deletions docker/dockerfile.merlin
View file
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
# syntax=docker/dockerfile:1.2
ARG TRITON_VERSION=22.08
ARG DLFW_VERSION=22.08
ARG TRITON_VERSION=22.11
ARG DLFW_VERSION=22.11

ARG FULL_IMAGE=nvcr.io/nvidia/tritonserver:${TRITON_VERSION}-py3
ARG BASE_IMAGE=nvcr.io/nvidia/tritonserver:${TRITON_VERSION}-py3-min
Expand All @@ -11,7 +11,8 @@ FROM ${DLFW_IMAGE} as dlfw
FROM ${BASE_IMAGE} as build

# Args
ARG DASK_VER=2022.05.1
ARG DASK_VER=2022.07.1
ARG MERLIN_VER=main
ARG CORE_VER=main
ARG MODELS_VER=main
ARG NVTAB_VER=main
Expand All @@ -20,12 +21,14 @@ ARG SYSTEMS_VER=main
ARG TF4REC_VER=main
ARG DL_VER=main

ENV MERLIN_VER=${MERLIN_VER}
ENV CORE_VER=${CORE_VER}
ENV MODELS_VER=${MODELS_VER}
ENV NVTAB_VER=${NVTAB_VER}
ENV NVTAB_BACKEND_VER=${NVTAB_BACKEND_VER}
ENV SYSTEMS_VER=${SYSTEMS_VER}
ENV TF4REC_VER=${TF4REC_VER}
ENV DL_VER=${DL_VER}

# Envs
ENV CUDA_HOME=/usr/local/cuda
Expand All @@ -36,7 +39,7 @@ ENV LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda/lib64:/usr/local/cuda/extra
ENV PATH=${CUDA_HOME}/lib64/:${PATH}:${CUDA_HOME}/bin

# Set up NVIDIA package repository
RUN apt update -y --fix-missing && \
RUN apt clean && apt update -y --fix-missing && \
apt install -y --no-install-recommends software-properties-common && \
wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2004/x86_64/cuda-ubuntu2004.pin && \
mv cuda-ubuntu2004.pin /etc/apt/preferences.d/cuda-repository-pin-600 && \
Expand Down Expand Up @@ -92,7 +95,7 @@ RUN pip install "cmake<3.25.0" ninja scikit-build

RUN pip install pandas==1.3.5
RUN pip install cupy-cuda117 nvidia-pyindex pybind11 pytest transformers==4.12 tensorflow-metadata
RUN pip install betterproto cachetools graphviz nvtx scipy scikit-learn
RUN pip install betterproto cachetools graphviz nvtx scipy "scikit-learn<1.2"
RUN pip install tritonclient[all] grpcio-channelz fiddle wandb npy-append-array
RUN pip install git+https://github.com/rapidsai/asvdb.git@main
# xgboost and treelite are pinned from tritonserver support
Expand All @@ -103,7 +106,6 @@ RUN pip install "cuda-python>=11.5,<12.0"
RUN pip install fsspec==2022.5.0 llvmlite
RUN pip install pynvml dask==${DASK_VER} distributed==${DASK_VER}
RUN pip install numpy==1.22.4

# protobuf versions
# tensorflow: >= 3.9.2, < 3.20
# The upper bound here is not strict (it will work with 3.20, though not on Windows, and not with 4.x)
Expand All @@ -116,7 +118,6 @@ RUN pip install numpy==1.22.4
# https://github.com/onnx/onnx/blob/v1.13.0/requirements.txt#L2
RUN pip install protobuf==3.20.3


# Triton Server
WORKDIR /opt/tritonserver
COPY --chown=1000:1000 --from=triton /opt/tritonserver/LICENSE .
Expand All @@ -135,18 +136,6 @@ ENV LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/opt/tritonserver/lib

RUN pip install dask==${DASK_VER} distributed==${DASK_VER} dask[dataframe]==${DASK_VER}

# Install faiss (with sm80 support since the faiss-gpu wheels
# don't include it https://github.com/kyamagu/faiss-wheels/issues/54)
RUN git clone --branch v1.7.2 https://github.com/facebookresearch/faiss.git build-env && \
pushd build-env && \
cmake -B build . -DFAISS_ENABLE_GPU=ON -DFAISS_ENABLE_PYTHON=ON -DBUILD_TESTING=OFF -DCMAKE_BUILD_TYPE=Release -DCMAKE_CUDA_ARCHITECTURES="60;70;80" && \
make -C build -j $(nproc) faiss swigfaiss && \
pushd build/faiss/python && \
python setup.py install && \
popd && \
popd && \
rm -rf build-env

# Install NVTabular Triton Backend
ARG TRITON_VERSION
RUN git clone https://github.com/NVIDIA-Merlin/nvtabular_triton_backend.git build-env && \
Expand All @@ -164,61 +153,25 @@ RUN git clone https://github.com/NVIDIA-Merlin/nvtabular_triton_backend.git buil
popd && \
rm -rf build-env

# Install spdlog
RUN git clone --branch v1.9.2 https://github.com/gabime/spdlog.git build-env && \
# Install faiss (with sm80 support since the faiss-gpu wheels
# don't include it https://github.com/kyamagu/faiss-wheels/issues/54)
RUN git clone --branch v1.7.2 https://github.com/facebookresearch/faiss.git build-env && \
pushd build-env && \
mkdir build && cd build && cmake .. && make -j && make install && \
cmake -B build . -DFAISS_ENABLE_GPU=ON -DFAISS_ENABLE_PYTHON=ON -DBUILD_TESTING=OFF -DCMAKE_BUILD_TYPE=Release -DCMAKE_CUDA_ARCHITECTURES="60;70;80" && \
make -C build -j $(nproc) faiss swigfaiss && \
pushd build/faiss/python && \
python setup.py install && \
popd && \
popd && \
rm -rf build-env

# Install arrow
ENV ARROW_HOME=/usr
RUN git clone --branch apache-arrow-7.0.0 --recurse-submodules https://github.com/apache/arrow.git build-env && \
# Install spdlog
RUN git clone --branch v1.9.2 https://github.com/gabime/spdlog.git build-env && \
pushd build-env && \
export PARQUET_TEST_DATA="${PWD}/cpp/submodules/parquet-testing/data" && \
export ARROW_TEST_DATA="${PWD}/testing/data" && \
pip install -r python/requirements-build.txt && \
mkdir cpp/release && \
pushd cpp/release && \
cmake -DCMAKE_INSTALL_PREFIX=${ARROW_HOME} \
-DCMAKE_INSTALL_LIBDIR=lib \
-DCMAKE_LIBRARY_PATH=${CUDA_CUDA_LIBRARY} \
-DARROW_FLIGHT=ON \
-DARROW_GANDIVA=OFF \
-DARROW_ORC=ON \
-DARROW_WITH_BZ2=ON \
-DARROW_WITH_ZLIB=ON \
-DARROW_WITH_ZSTD=ON \
-DARROW_WITH_LZ4=ON \
-DARROW_WITH_SNAPPY=ON \
-DARROW_WITH_BROTLI=ON \
-DARROW_PARQUET=ON \
-DARROW_PYTHON=ON \
-DARROW_PLASMA=ON \
-DARROW_BUILD_TESTS=ON \
-DARROW_CUDA=ON \
-DARROW_DATASET=ON \
-DARROW_HDFS=ON \
-DARROW_S3=ON \
.. && \
make -j$(nproc) && \
make install && \
popd && \
pushd python && \
export PYARROW_WITH_PARQUET=ON && \
export PYARROW_WITH_CUDA=ON && \
export PYARROW_WITH_ORC=ON && \
export PYARROW_WITH_DATASET=ON && \
export PYARROW_WITH_S3=ON && \
export PYARROW_WITH_HDFS=ON && \
python setup.py build_ext --build-type=release bdist_wheel && \
pip install dist/*.whl --no-deps --force-reinstall && \
popd && \
mkdir build && cd build && cmake .. && make -j && make install && \
popd && \
rm -rf build-env



# Clean up
RUN rm -rf /repos

Expand Down Expand Up @@ -284,7 +237,7 @@ COPY --chown=1000:1000 --from=build /usr/include /usr/include/
COPY --chown=1000:1000 --from=build /usr/local/include /usr/local/include/
COPY --chown=1000:1000 --from=build /usr/lib/ /usr/lib/
COPY --chown=1000:1000 --from=build /usr/local/lib/ /usr/local/lib/

COPY --chown=1000:1000 --from=triton /usr/lib/x86_64-linux-gnu/libdcgm.so.2 /usr/lib/x86_64-linux-gnu/libdcgm.so.2

# Binaries
COPY --chown=1000:1000 --from=build /usr/local/bin /usr/local/bin/
Expand All @@ -302,6 +255,9 @@ COPY --chown=1000:1000 --from=triton /opt/tritonserver/backends/python backends/
COPY --chown=1000:1000 --from=triton /opt/tritonserver/backends/fil backends/fil/
COPY --chown=1000:1000 --from=triton /usr/bin/serve /usr/bin/.
COPY --chown=1000:1000 --from=triton /usr/lib/x86_64-linux-gnu/libdcgm.so.2 /usr/lib/x86_64-linux-gnu/libdcgm.so.2
COPY --chown=1000:1000 --from=triton /usr/local/cuda-11.8/targets/x86_64-linux/lib/libcupti.so.11.8 /usr/local/cuda-11.8/targets/x86_64-linux/lib/libcupti.so.11.8



ENV PATH=/opt/tritonserver/bin:${PATH}:
ENV LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/opt/tritonserver/lib
Expand All @@ -317,23 +273,33 @@ ENV PYTHONPATH=$PYTHONPATH:/usr/local/lib/python3.8/dist-packages/

# rapids components from the DLFW image
COPY --chown=1000:1000 --from=dlfw /usr/lib/libcudf* /usr/lib/
COPY --chown=1000:1000 --from=dlfw /usr/lib/libarrow* /usr/lib/
COPY --chown=1000:1000 --from=dlfw /usr/lib/libparquet* /usr/lib/
COPY --chown=1000:1000 --from=dlfw /usr/lib/libnvcomp* /usr/lib/
COPY --chown=1000:1000 --from=dlfw /usr/include/cudf /usr/include/cudf/
COPY --chown=1000:1000 --from=dlfw /usr/include/rmm /usr/include/rmm/
ARG PYTHON_VERSION=3.8
COPY --chown=1000:1000 --from=dlfw /usr/local/lib/python${PYTHON_VERSION}/dist-packages/rmm /usr/local/lib/python${PYTHON_VERSION}/dist-packages/rmm
COPY --chown=1000:1000 --from=dlfw /usr/local/lib/python${PYTHON_VERSION}/dist-packages/cuda /usr/local/lib/python${PYTHON_VERSION}/dist-packages/cuda
COPY --chown=1000:1000 --from=dlfw /usr/local/lib/python${PYTHON_VERSION}/dist-packages/pyarrow /usr/local/lib/python${PYTHON_VERSION}/dist-packages/pyarrow
COPY --chown=1000:1000 --from=dlfw /usr/local/lib/python${PYTHON_VERSION}/dist-packages/cudf /usr/local/lib/python${PYTHON_VERSION}/dist-packages/cudf
COPY --chown=1000:1000 --from=dlfw /usr/local/lib/python${PYTHON_VERSION}/dist-packages/dask_cudf /usr/local/lib/python${PYTHON_VERSION}/dist-packages/dask_cudf
COPY --chown=1000:1000 --from=dlfw /usr/local/lib/python${PYTHON_VERSION}/dist-packages/dask_cuda /usr/local/lib/python${PYTHON_VERSION}/dist-packages/dask_cuda

COPY --chown=1000:1000 --from=dlfw /usr/local/lib/python3.8/dist-packages/cudf-*.dist-info /usr/local/lib/python3.8/dist-packages/cudf.dist-info/
COPY --chown=1000:1000 --from=dlfw /usr/local/lib/python3.8/dist-packages/dask_cudf-*.dist-info /usr/local/lib/python3.8/dist-packages/dask_cudf.dist-info/
COPY --chown=1000:1000 --from=dlfw /usr/local/lib/python3.8/dist-packages/dask_cuda-*.dist-info /usr/local/lib/python3.8/dist-packages/dask_cuda.dist-info/
COPY --chown=1000:1000 --from=dlfw /usr/local/lib/python3.8/dist-packages/pyarrow-*.dist-info /usr/local/lib/python3.8/dist-packages/pyarrow.dist-info/
COPY --chown=1000:1000 --from=dlfw /usr/local/lib/python3.8/dist-packages/rmm-*.dist-info /usr/local/lib/python3.8/dist-packages/rmm.dist-info/


RUN pip install --no-cache-dir jupyterlab pydot testbook numpy==1.22.4

ENV JUPYTER_CONFIG_DIR=/tmp/.jupyter
ENV JUPYTER_DATA_DIR=/tmp/.jupyter
ENV JUPYTER_RUNTIME_DIR=/tmp/.jupyter

ARG MERLIN_VER
ARG CORE_VER
ARG MODELS_VER
ARG NVTAB_VER
Expand All @@ -342,7 +308,8 @@ ARG TF4REC_VER
ARG DL_VER

# Add Merlin Repo
RUN git clone https://github.com/NVIDIA-Merlin/Merlin/ /Merlin
RUN git clone https://github.com/NVIDIA-Merlin/Merlin/ /Merlin && \
cd /Merlin/ && git checkout ${MERLIN_VER} && pip install . --no-deps

# Install Merlin Core
RUN git clone https://github.com/NVIDIA-Merlin/core.git /core/ && \
Expand All @@ -352,7 +319,6 @@ RUN git clone https://github.com/NVIDIA-Merlin/core.git /core/ && \
RUN git clone https://github.com/NVIDIA-Merlin/dataloader.git /dataloader/ && \
cd /dataloader/ && git checkout ${DL_VER} && pip install . --no-deps


# Install NVTabular
ENV PROTOCOL_BUFFERS_PYTHON_IMPLEMENTATION='python'
RUN git clone https://github.com/NVIDIA-Merlin/NVTabular.git /nvtabular/ && \
Expand Down

0 comments on commit efa8215

Please sign in to comment.