README.md

Bonus Challenge! Deploy your own facial detection alogoritm in SageMaker

In this lab session, you will learn how to build and train a model in Amazon SageMaker

Before we begin, download the repo to your local directory. You can download the repo from root directory

Click Clone or download button on the right.

This will download the repo as a zip file. Extract the zip file.

Step 1- Create notebook instance

Go to SageMaker console: https://console.aws.amazon.com/sagemaker

Click on Create Notebook instance

Step 2- Notebook specifications

1. Provide the name of the notebook as face-detection
2. Notebook instance type- ml.t2.medium
3. IAM role- Click 'Create a new role'

In the dialog box that opens up:

1. Click 'Any S3 bucket'
2. Leave the rest as is. Click 'Create role'

Leave VPC, Lifecycle and Encryption key as defaults. Dont make any changes

Click 'Create notebook instance'

Step 3- View notebook instances

You can view all your notebook instances by choosing Notebook on the left menu. It will take couple of minutes for the notebook instance to be created.

Step 4- Upload and Open notebook

Choose Upload button on the jupyter page

Find the SSD_Object_Detection_SageMaker_v3.ipynb file (You can find it in the sagemaker lab directory of the extracted repo. You downloaded and extracted the zip file earlier in the process)

Choose Upload

You can choose your uploaded notebook and click on 'Open'.

This will open your Jupyter notebook.

Step 5- Execute notebook

1. Execute the cells by clicking on run button or using shift+ enter on your keyboard

Congratulations! You have successfully deployed your own face detection model with Amazon Sagemaker! Next, follow the same steps in the previous challenges to deploy your project and lambda function -- except this time, use your created SageMaker model instead of the one provided by default with AWS Deep Lens.

---

Back (Challenge 2: Sentiment Analysis) | Next (Cleanup Your Resources)