This instruction provides a starting point for a ROCm installation of hiptensorflow (mostly via deb packages). Note: it is recommended to start with a clean Ubuntu 16.04 system
export ROCM_PATH=/opt/rocm
export DEBIAN_FRONTEND noninteractive
sudo apt update && sudo apt install -y wget software-properties-common
Add the ROCm repository:
wget -qO - http://repo.radeon.com/rocm/apt/debian/rocm.gpg.key | sudo apt-key add -
sudo sh -c 'echo deb [arch=amd64] http://repo.radeon.com/rocm/apt/debian/ xenial main > /etc/apt/sources.list.d/rocm.list'
Install misc pkgs:
sudo apt-get update && sudo apt-get install -y \
build-essential \
clang-3.8 \
clang-format-3.8 \
clang-tidy-3.8 \
cmake \
cmake-qt-gui \
ssh \
curl \
apt-utils \
pkg-config \
g++-multilib \
git \
libunwind-dev \
libfftw3-dev \
libelf-dev \
libncurses5-dev \
libpthread-stubs0-dev \
vim \
gfortran \
libboost-program-options-dev \
libssl-dev \
libboost-dev \
libboost-system-dev \
libboost-filesystem-dev \
rpm \
wget && \
sudo apt-get clean && \
sudo rm -rf /var/lib/apt/lists/*
Install ROCm pkgs:
sudo apt-get update && \
sudo apt-get install -y --allow-unauthenticated \
rocm-dkms rocm-dev rocm-libs \
rocm-device-libs \
hsa-ext-rocr-dev hsakmt-roct-dev hsa-rocr-dev \
rocm-opencl rocm-opencl-dev \
rocm-utils \
rocm-profiler cxlactivitylogger \
miopen-hip miopengemm
Setup environment variables, and add those environment variables at the end of ~/.bashrc
export HCC_HOME=/opt/rocm/hcc
export HIP_PATH=/opt/rocm/hip
export PATH=$HCC_HOME/bin:$HIP_PATH/bin:$PATH
git clone -b develop https://github.com/ROCmSoftwarePlatform/hipeigen.git /opt/rocm/hipeigen
sudo apt-get update && sudo apt-get install -y \
python-numpy \
python-dev \
python-wheel \
python-mock \
python-future \
python-pip \
python-yaml \
python-setuptools && \
sudo apt-get clean && \
sudo rm -rf /var/lib/apt/lists/*
cd ~ && git clone https://github.com/soumith/convnet-benchmarks.git
cd ~ && git clone https://github.com/tensorflow/models.git
# Wget the hipTF whl package
cd ~/ && wget http://repo.radeon.com/rocm/misc/tensorflow/tensorflow-1.0.1-cp27-cp27mu-linux_x86_64.whl
# Pip install the whl package
cd ~ && sudo pip install tensorflow-1.0.1-cp27-cp27mu-linux_x86_64.whl && rm tensorflow-1.0.1-cp27-cp27mu-linux_x86_64.whl