This project implements a web application for sentiment analysis using Amazon SageMaker, AWS API Gateway, AWS Lambda, PyTorch, Scikit-learn, and pickle. The app utilizes a recurrent neural network (RNN/LSTM) model trained on movie reviews from the iMDb dataset to predict sentiment. Prior familiarity with SageMaker, especially completion of the mini-project "Sentiment Analysis using XGBoost," was crucial to understanding the main SageMaker resources required for this project.
The goal was to construct a complete end-to-end project, enabling users to input a movie review through a simple web page. The review is then sent to the deployed model for sentiment prediction. Template code is provided, and additional functionality was implemented as required, following specific instructions.
You can find a detailed description of how this project was made in SageMaker Project.ipynb.
Every code after #TODO comments were written by Rodolfo Teles.
- AWS SageMaker
- AWS API Gateway
- AWS Lambda
- sagemaker
- pytorch
- scikit-learn
- numpy
- pickle
- pandas
- boto3
Please see the README in the root directory of the Udacity GitHub Page for instructions on setting up a SageMaker notebook and downloading the project files (as well as the other notebooks).