setup.sh

#!/bin/bash

if [ -d "$1" ]; then
  # Take action if $DIR exists. #
  echo "Using openvino source $1..."
else
  echo "Please provide openvino source DIR ."
  exit 1
fi

# openvino source code root
ov_src=`realpath $1`

ln -sf $ov_src openvino

mkdir -p extern

cd extern

mkdir -p cv_bench_cache
sudo mount -t nfs -o ro 10.91.242.212:/data/cv_bench_cache ./cv_bench_cache

mkdir -p omz-validation-datasets
#sudo mount -t nfs -o ro 10.91.242.212:/data/nn_icv_cv_externalN/omz-validation-datasets ./omz-validation-datasets
sudo mount -t nfs -o ro ov-share-02.iotg.sclab.intel.com:/data/nn_icv_cv_externalN/omz-validation-datasets/ ./omz-validation-datasets

#mkdir nfs_share_data
#sudo mount -t nfs -o ro 10.67.107.130:/home/share nfs_share_data

#mkdir nfs_share_models
#sudo mount -t nfs 10.67.108.173:/home/vsi/nfs_share nfs_share_models

ln -sf `realpath ./cv_bench_cache/ww06_weekly_23.0-9585-ab509ce1645-API2.0_int8` ww06_weekly_23.0-9585-ab509ce1645-API2.0_int8
ln -sf `realpath ./cv_bench_cache/ww09_weekly_23.0.0-9828-4fd38844a28-API2.0` ww09_weekly_23.0.0-9828-4fd38844a28-API2.0


python3 -m virtualenv ./venv
. ./venv/bin/activate

pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cpu
pip install -r ../requirements.txt
## install accuracy_checker in-place
pip install -e $ov_src/thirdparty/open_model_zoo/tools/accuracy_checker/

mkdir -p annotations

# complementary repo that provides more annotation converters
git clone git@gitlab-icv.toolbox.iotg.sclab.intel.com:algo/annotation_converters_ext.git &&
cd annotation_converters_ext &&
python3 setup.py install_as_extension --accuracy-checker-dir=../../openvino/thirdparty/open_model_zoo/tools/accuracy_checker

git clone git@github.com:intel-innersource/frameworks.ai.openvino.accuracy-checker-configs.git