predict_custom_trained_model_sample function should end with return(predictions)

#### Environment details

- OS type and version: Windows 10
- Python version: 3.7
- pip version: 20.3.4
- `google-cloud-aiplatform` version: `pip show google-cloud-aiplatform`

#### Steps to reproduce

1. after importing a Keras model to Vertex AI and deploying it to an endpoint, I follow link to **Sample Request** that gives details on how to get a prediction from the endpoint https://screenshot.googleplex.com/4RRUZXwXmgwjpaN
2. When you follow that **Sample Request** link, you get this page https://screenshot.googleplex.com/98PGUjY9Co3ujoh. Step 2 from these instructions points to https://github.com/googleapis/python-aiplatform/blob/main/samples/snippets/prediction_service/predict_custom_trained_model_sample.py, which includes the definition for `predict_custom_trained_model_sample` function
3. Following those instructions, I create a Flask server module and copy in the definition of `predict_custom_trained_model_sample` function from https://github.com/googleapis/python-aiplatform/blob/main/samples/snippets/prediction_service/predict_custom_trained_model_sample.py to call the model I deployed to a Vertex AI endpoint. Code for this module is here: https://github.com/ryanmark1867/deep_learning_web_deployment/blob/master/flask_endpoint_deploy.py
4. To get results back from the `predict_custom_trained_model_sample` function, I needed to update it to include a return statement: `return(predictions)`
5. Net: the definition of `predict_custom_trained_model_sample` function at https://github.com/googleapis/python-aiplatform/blob/main/samples/snippets/prediction_service/predict_custom_trained_model_sample.py needs to be updated to include a `return(predictions)` statement at the end.

#### Code example

```python
# example
```

#### Stack trace
```
# example
```

Making sure to follow these steps will guarantee the quickest resolution possible.

Thanks!