Openvino Workshop

This workshope contains jupyter notebook which describes full OpenVINO workflow. In this workshop you will try most componrnts of the Openvino Toolkit:

1. Model Downloader - tool that downloads models from Open ModelZoo
2. Open ModelZoo - set of pre-trained and open-source models that you can use for learning and demo purposes or for developing deep learning software.
3. Model Optimizer - tool that facilitates the transition between the training and deployment environment, performs static model analysis, and adjusts deep learning models for optimal execution on end-point target devices.
4. Inference Engine - a unified API to allow high performance inference on many hardware.
5. Accuracy Checker - tool to measure accuracy of a model.
6. Post-Training Optimization Toolkit - tool designed to convert a model into a more hardware-friendly representation by applying specific methods that do not require re-training, for example, post-training quantization.
7. Benchmark Application - C++ sample for getting best performance with the OpenVINO Inference Engine.
8. Python API for Inference Engine

How to run the workshop

From Docker

You can find docker image with this notebook by following link: https://hub.docker.com/r/artyomtugaryov/openvino_workshop

To tun this docker use the following link:

    docker run -p 8888:8888 -it artyomtugaryov/openvino_workshop

From sources

1. Install OpenVINO

2. Run a bootstrap script to install all dependencies and run the jupyter notebook with the workshop:

    ./bootstrap/bootstrap.sh -start

In the successful case you will see the following message:

    To access the notebook, open this file in a browser:
        file:/.../jupyter/runtime/nbserver-18647-open.html
    Or copy and paste one of these URLs:
        http://localhost:8888/?token=...
     or http://127.0.0.1:8888/?token=...
    Opening in existing browser session.

If you get any error, you can setup dependencies manually:

3. Open a terminal and activate OpenVINO environment:

    source ~/intel/openvino/bin/setupvars.sh

4. Clone the repository:

    git clone https://github.com/artyomtugaryov/openvino_workshop.git

5. Change directory to workshop root path:

    cd openvino_workshop

6. Create Virtual environment for python3:

    python3 -m pip install virtualenv
    
    python3 -m virtualenv -p `which python3` env

    source env/bin/activate

7. Install workshop dependencies:

    pip install -r requirements.txt

8. Install OpenVINO python dependencies:

   8.1 Install Accuracy Checker

       pushd ${INTEL_OPENVINO_DIR}/deployment_tools/open_model_zoo/tools/accuracy_checker/
           python setup.py install
       popd

   8.2 Install Post-Training Optimization Toolkit

       pushd ${INTEL_OPENVINO_DIR}/deployment_tools/tools/post_training_optimization_toolkit/
           python setup.py install
       popd

   8.3 Install dependencies for the Model Downloader

       python -m pip install -r ${INTEL_OPENVINO_DIR}/deployment_tools/open_model_zoo/tools/downloader/requirements.in

   8.4 Install dependencies for the Model Optimizer

       python -m pip install -r ${INTEL_OPENVINO_DIR}/deployment_tools/model_optimizer/requirements.txt

9. Install gettext-base package for using envsubst:

    apt-get install gettext-base

10. Run the notebook:

   jupyter notebook