Docker.md

Digit recognition model with MNIST dataset using Docker container

In this example, we show how to use a pre-trained custom MNIST model to performing real time Digit recognition with TorchServe. We will be serving the model using a Docker container.

The inference service would return the digit inferred by the model in the input image.

We used the following pytorch example to train the basic MNIST model for digit recognition : https://github.com/pytorch/examples/tree/master/mnist

Serve an MNIST model on TorchServe docker container

Run the commands given in following steps from the parent directory of the root of the repository. For example, if you cloned the repository into /home/my_path/serve, run the steps from /home/my_path/serve

Create a torch model archive using the torch-model-archiver utility to archive the above files.

torch-model-archiver --model-name mnist --version 1.0 --model-file examples/image_classifier/mnist/mnist.py --serialized-file examples/image_classifier/mnist/mnist_cnn.pt --handler  examples/image_classifier/mnist/mnist_handler.py

Move .mar file into model_store directory

mkdir model_store
mv mnist.mar model_store/

Start a docker container with torchserve

docker run --rm -it -p 127.0.0.1:8080:8080 -p 127.0.0.1:8081:8081 -p 127.0.0.1:8082:8082 -v $(pwd)/model_store:/home/model-server/model-store pytorch/torchserve:latest-cpu

Register the model on TorchServe using the above model archive file

curl -X POST "localhost:8081/models?model_name=mnist&url=mnist.mar&initial_workers=4"

If this succeeeds, you will see a message like below

{
"status": "Model \"mnist\" Version: 1.0 registered with 4 initial workers"
}

Run digit recognition inference outside the container

curl http://127.0.0.1:8080/predictions/mnist -T examples/image_classifier/mnist/test_data/0.png

The output in this case will be a 0