Machine Learning Serving

Service Context Understanding with KFood DB

---

Required Software

https://github.com/imadelh/ML-web-app

https://github.com/aai4r/aai4r-ServiceContextUnderstanding

Install

1. Clone the Repo

git clone https://github.com/yunsujeon/MLserving_ServiceContextUnderstanding.git

2. Download model and locate at app/output/
   model_best.pth.tar
   faster_rcnn_1_7_9999.pth
   class_info_Kfood.pkl
   class_info_FoodX251.pkl
   class_info_Food101.pkl

Running on Local machine with Anaconda

1. Anaconda create and activate

conda create -n <name> python==3.6.2
conda activate <name>

2. Install requirements

pip install -r requirements.txt

3. Run

python app.py

Go to http://0.0.0.0:8888 , then you can see wep page and explanation.

Running on Docker

1. Install Docker your self

2. Create Docker image by build Dockerfile

sudo docker build -t <image name> .
or
docker build -t <image name> .

3. Run docker file

docker run -i -t --rm -p 8888:8888 -v <your path>:/<docker path> --shm-size=2GB --gpus all <image name>
ex)
docker run -i -t --rm -p 8888:8888 -v /home/intern/MLserving/app:/app --shm-size=2GB --gpus all <image name>

If you need more memory in docker env, and select specific gpus ..
--shm-size=8G
--gpus '"device=0,1"'

Go to http://0.0.0.0:8888, then you can see wep page and explanation.

Improvement

You can run this codes at SSH server, Its all same this repo's local, docker examples

You have to match CUDA version to SSH server. So you have to change Dockerfile

But you will change the access url
0.0.0:8888 -> [your remote server ip]:8888

Enjoy this Repo. thank you.