# TGL Sales Kit with OpenVINO 2021.4 LTS Release
###### tags: `Intel AI Sales Kit`
- Table of Content
[ToC]
### 1. Prepare docker image patch
#### Get required files from openvino/ubuntu20_data_dev:2021.4_tgl
```bash=
docker run -it -v /tmp/.X11-unix:/tmp/.X11-unix -e DISPLAY=$DISPLAY -v ~/Downloads:/mnt --device /dev/dri:/dev/dri --group-add=$(stat -c "%g" /dev/dri/render*) -u root --rm openvino/ubuntu20_data_dev:2021.4_tgl
mkdir -p /mnt/openvino_2021.4.582/python/python3.8/openvino/
cp -ar /opt/intel/openvino_2021.4.582/python/python3.8/openvino/tools /mnt/openvino_2021.4.582/python/python3.8/openvino/
mkdir -p /mnt/openvino_2021.4.582/deployment_tools/tools/benchmark_tool/
cp -ar /opt/intel/openvino_2021.4.582/deployment_tools/tools/benchmark_tool/* /mnt/openvino_2021.4.582/deployment_tools/tools/benchmark_tool/
mkdir -p /mnt/openvino_2021.4.582/data_processing/dl_streamer/samples/
cp -ar /opt/intel/openvino_2021.4.582/data_processing/dl_streamer/samples/model_proc/ /mnt/openvino_2021.4.582/data_processing/dl_streamer/samples/
exit
```
#### Prepare YOLO v3/v4/v5l FP32/FP16/FP16-INT8 model
YOLO v3 : https://hackmd.io/lKCVqs5xQ3WSFvl33wLnjQ
YOLO v4 : https://hackmd.io/mQHIp8vgTViF9Ao4mnI5rw
YOLO v5l : https://hackmd.io/EwcxDpOETz-Ujz6q6pFWXQ
At last steps of above three model conversion, in the container, copy /home/openvino/openvino_models foler to /mnt folder. They will be accessable in ~/Downloads folder in the host
```=bash
cd ~/Downloads
tar -czf openvino_2021.4.582_saleskit.tgz openvino_2021.4.582
tar -czf openvino_models.tgz openvino_models
```
### 2. Build docker image
```bash=
cd
mkdir tgl_saleskit_2021.4_tgl
cd tgl_saleskit_2021.4_tgl
cp ~/Downloads/penvino_2021.4.582_saleskit.tgz .
cp ~/Downloads/openvino_models.tgz .
# Add smartcity_demo.sh here
docker build --build-arg models_url=openvino_models.tgz --build-arg patch_url=openvino_2021.4.582_saleskit.tgz -t tgl_saleskit:2021.4_tgl .
```
#### smartcity_demo.sh (2021.4_tgl)
```
#!/bin/bash
source /opt/intel/openvino/bin/setupvars.sh
cd
gst-launch-1.0 filesrc location=/mnt/NewVideo2.mp4 ! decodebin ! gvadetect model=/opt/intel/openvino_models/public/yolo-v3-tf/FP16-INT8/yolo-v3-tf.xml model-proc=/opt/intel/openvino_2021.4.582/data_processing/dl_streamer/samples/model_proc/public/object_detection/yolo-v3-tf.json device=GPU ! queue ! gvawatermark ! videoconvert ! fpsdisplaysink video-sink=xvimagesink sync=false
```
#### Dockerfile
```dockerfile=
# Copyright (C) 2019-2021 Intel Corporation
# SPDX-License-Identifier: Apache-2.0
FROM openvino/ubuntu20_data_runtime:2021.4_tgl AS base
USER openvino
WORKDIR /opt/intel/openvino_2021.4.582
SHELL ["/bin/bash", "-xo", "pipefail", "-c"]
ARG patch_url
ADD ${patch_url} /opt/intel
ARG models_url
ADD ${models_url} /opt/intel
# setup Python
ENV PYTHON_VER python3.8
RUN ${PYTHON_VER} -m pip install --no-cache-dir -r ${INTEL_OPENVINO_DIR}/python/${PYTHON_VER}/requirements.txt && \
${PYTHON_VER} -m pip install --no-cache-dir -r ${INTEL_OPENVINO_DIR}/deployment_tools/tools/benchmark_tool/requirements.txt;
ADD smartcity_demo.sh /home/openvino
CMD ["/bin/bash"]
```
### 3. Upload docker image to dockerhub
#### 3.1 docker login
```
docker login
Login with your Docker ID to push and pull images from Docker Hub. If you don't have a Docker ID, head over to https://hub.docker.com to create one.
Username: your_dockerhub_id
Password:
WARNING! Your password will be stored unencrypted in /home/openvino/.docker/config.json.
Configure a credential helper to remove this warning. See
https://docs.docker.com/engine/reference/commandline/login/#credentials-store
Login Succeeded
```
#### 3.2 docker push (uploade docker image to dockerhub)
```
docker tag tgl_saleskit:2021.4_tgl your_dockerhub_id/tgl_saleskit:2021.4_tgl
docker push your_dockerhub_id/tgl_saleskit:2021.4_tgl
```
##### logs
```bash=
dlcv@dlcv-NUC11PAHi7:~/tgl_saleskit_2021.4_tgl$ docker images
REPOSITORY TAG IMAGE ID CREATED SIZE
tgl_saleskit 2021.4_tgl 292fb6a5c80c 17 minutes ago 3.01GB
openvino/ubuntu20_data_dev 2021.4_tgl 2844df1d1032 5 weeks ago 6.72GB
openvino/ubuntu20_data_runtime 2021.4_tgl e42518d8e4c0 5 weeks ago 1.98GB
your_dockerhub_id/openvino 2021.3_developer_models 3a1328523841 3 months ago 3.29GB
dlcv@dlcv-NUC11PAHi7:~/tgl_saleskit_2021.4_tgl$ docker tag tgl_saleskit:2021.4_tgl your_dockerhub_id/tgl_saleskit:2021.4_tgl
dlcv@dlcv-NUC11PAHi7:~/tgl_saleskit_2021.4_tgl$ docker images
REPOSITORY TAG IMAGE ID CREATED SIZE
your_dockerhub_id/tgl_saleskit 2021.4_tgl 292fb6a5c80c 19 minutes ago 3.01GB
tgl_saleskit 2021.4_tgl 292fb6a5c80c 19 minutes ago 3.01GB
openvino/ubuntu20_data_dev 2021.4_tgl 2844df1d1032 5 weeks ago 6.72GB
openvino/ubuntu20_data_runtime 2021.4_tgl e42518d8e4c0 5 weeks ago 1.98GB
your_dockerhub_id/openvino 2021.3_developer_models 3a1328523841 3 months ago 3.29GB
```
### 4. Run Smart City Demo
```bash=
wget -O ~/Downloads/NewVideo2.mp4 https://github.com/incluit/OpenVino-For-SmartCity/raw/master/data/NewVideo2.mp4
docker pull your_dockerhub_id/tgl_saleskit:2021.4_tgl
docker run -it -v /tmp/.X11-unix:/tmp/.X11-unix -e DISPLAY=$DISPLAY -v ~/Downloads:/mnt --device /dev/dri:/dev/dri --group-add=$(stat -c "%g" /dev/dri/render*) --rm your_dockerhub_id/tgl_saleskit:2021.4_tgl /home/openvino/smartcity_demo.sh
```
### 5. Run Benchmark
#### YOLO v3 416x416 : Throughput: 59 FPS
```bash=
docker run -it -v /tmp/.X11-unix:/tmp/.X11-unix -e DISPLAY=$DISPLAY -v ~/Downloads:/mnt --device /dev/dri:/dev/dri --group-add=$(stat -c "%g" /dev/dri/render*) --rm your_dockerhub_id/tgl_saleskit:2021.4_tgl
python3 /opt/intel/openvino/deployment_tools/tools/benchmark_tool/benchmark_app.py -m /opt/intel/openvino_models/public/yolo-v3-tf/FP16-INT8/yolo-v3-tf.xml -d GPU -api async -t 60
```
#### YOLO V4 vs YOLO v5-l @ 640x640 input resolution
YOLO v4 640x640 : Throughput: 23.12 FPS
```bash=
docker run -it -v /tmp/.X11-unix:/tmp/.X11-unix -e DISPLAY=$DISPLAY -v ~/Downloads:/mnt --device /dev/dri:/dev/dri --group-add=$(stat -c "%g" /dev/dri/render*) --rm your_dockerhub_id/tgl_saleskit:2021.4_tgl
python3 /opt/intel/openvino/deployment_tools/tools/benchmark_tool/benchmark_app.py -m /mnt/openvino_models/public/yolo-v4-tf/FP16-INT8/yolo-v4-tf.xml -d GPU -shape [1,3,640,640]
```
YOLO v5l 640x640 : Throughput: 30.59 FPS
```bash=
docker run -it -v /tmp/.X11-unix:/tmp/.X11-unix -e DISPLAY=$DISPLAY -v ~/Downloads:/mnt --device /dev/dri:/dev/dri --group-add=$(stat -c "%g" /dev/dri/render*) --rm your_dockerhub_id/tgl_saleskit:2021.4_tgl
python3 /opt/intel/openvino/deployment_tools/tools/benchmark_tool/benchmark_app.py -m /mnt/openvino_models/public/yolo-v5l-v5-pt/FP16-INT8/yolo_v5l-v5.xml -d GPU
```
### 6. Add New Models for Benchmark
Follow steps in link below to convert models to FP32, FP16 and INT8 IR in /home/openvino/openvino_models. Then copy it to /mnt folder. They will be available in ~/Downloads folder in the host.
https://hackmd.io/3jNqTlEMTZe93rAKOpyMDA
### Appendix. Install OS and tool (skip this if your system have Ubuntu 20.04.2 or later version installed)
#### Make sure BIOS setting is correct
Please refer to
https://hackmd.io/UZ8LWR30Qn-SFPbC21_j5Q
#### Download and Install Ubuntu 20.04.2.0 LTS
URL to donwload Ubuntu 20.04.2.0 LTS ISO image:
https://ubuntu.com/download/desktop/thank-you?version=20.04.2.0&architecture=amd64
URL to donwload Ubuntu 20.04.3 LTS ISO image:
https://ubuntu.com/download/desktop/thank-you?version=20.04.3&architecture=amd64
#### Install Ubuntu 20.04
https://phoenixnap.com/kb/install-ubuntu-20-04
#### Install Docker Utility
```
sudo apt update
sudo apt-get remove docker docker-engine docker.io containerd runc
sudo apt install curl
curl -fsSL https://get.docker.com -o get-docker.sh
sudo sh get-docker.sh
sudo usermod -aG docker $USER
## Need to logout or reboot to run docker as non-root user
docker run hello-world
```
If you see error messages below, reboot your system.
```
openvino@openvino-TGL:~$ docker run hello-world
docker: Got permission denied while trying to connect to the Docker daemon socket at unix:///var/run/docker.sock: Post http://%2Fvar%2Frun%2Fdocker.sock/v1.24/containers/create: dial unix /var/run/docker.sock: connect: permission denied.
See 'docker run --help'.
```
Sign in with Wallet
Connect another wallet