![]() For the role, creating a new role works fine. The notebook name can be anything and using ml.t2.medium is a good idea as it is covered under the free tier. Log in to the AWS console and go to the SageMaker dashboard. Log in to the AWS console and create a notebook instance ![]() The notebook in this repository is intended to be executed using Amazon's SageMaker platform and the following is a brief set of instructions on setting up a managed notebook instance using SageMaker. After the predictor performs inference on the given input, it is returned to the web application via the Lambda function. The diagram below illustrates the flow of input from the web application to the model hosted on AWS, which is called using a Lambda function via a REST API. a larger training set, better tuned hyperparameters, or different ratios of training/validation/testing datasets).Īnother component of this project was an online form that can be used to pass inputs to the model to test it out. This could be due to a number of factors and it's quite possible that the neural net would perform better under different conditions (i.e. Prior to training this neural net, I also trained an XGBoost model, mostly just to see which one of the two models would perform better - the resulting scores were not sufficiently different that I could draw any meaningful conclusions, but XGBoost did slightly outperform the neural net. Everything for this project was done on Amazon Web Services (AWS) and their SageMaker platform as the goal of this project was to familiarize myself with the AWS ecosystem.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |