Improve docker image UX and fix MNIST serving example

docker/config.properties

@@ -3,6 +3,7 @@ management_address=http://0.0.0.0:8081
 metrics_address=http://0.0.0.0:8082
 grpc_inference_address=0.0.0.0
 grpc_management_address=0.0.0.0
+enable_envvars_config=true
 number_of_netty_threads=32
 job_queue_size=1000
 model_store=/home/model-server/model-store

docker/dockerd-entrypoint.sh

@@ -4,10 +4,10 @@ set -e
 
 if [[ "$1" = "serve" ]]; then
     shift 1
-    torchserve --start --ts-config /home/model-server/config.properties --disable-token-auth
+    torchserve --foreground --ts-config /home/model-server/config.properties --disable-token-auth "$@"
 else
     eval "$@"
-fi
 
-# prevent docker exit
-tail -f /dev/null
+    # prevent docker exit
+    tail -f /dev/null
+fi

examples/image_classifier/mnist/Docker.md

@@ -28,7 +28,7 @@ Run the commands given in following steps from the parent directory of the root
   ### Start a docker container with torchserve
 
   ```bash
-  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
+  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 -e TS_ENABLE_MODEL_API=true -v $(pwd)/model_store:/home/model-server/model-store pytorch/torchserve:latest-cpu
   ```
 
   ### Register the model on TorchServe using the above model archive file
@@ -45,6 +45,14 @@ Run the commands given in following steps from the parent directory of the root
   }
   ```
 
+  An alternative to manual registration of models is to specify the model names TorchServe should register at startup using the [`load_models`](https://pytorch.org/serve/configuration.html#load-models-at-startup) property. The property can be configured by setting the `TS_LOAD_MODELS=mnist.mar` environment variable (this removes the need for the `TS_ENABLE_MODEL_API` environment variable and the `curl` call above):
+
+  ```bash
+  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 -e TS_LOAD_MODELS=mnist.mar -v $(pwd)/model_store:/home/model-server/model-store pytorch/torchserve:latest-cpu
+  ```
+
+  Note that this approach does not allow specifying the initial number of workers.
+
   ### Run digit recognition inference outside the container
 
   ```bash