This is a sentiment analysis model trained on IMDB movies textual reviews. The resulting model will take as input a review from a simple web form and give a positive or negative output response.
The project is based on the following components:
- A PyTorch based Deep Neural Network model to predict sentiment from a review,
- An AWS Lambda server to execute on demand a code snippet that will provide data to the model endpoint,
- AWS API Gateway to gather data from a web page through a restful API **,
- A simple web page to input data. Two models were designed in the project, the second one is intended to improve initial performance.
Description/implementation of above items marked with ** is not part of the project repository.
NegativeReviewExample.png
PositiveReviewExample.png
SageMaker Project.ipynb
Web App Diagram.svg
sentiment_lambda_function.py
serve/model.py
serve/predict.py
serve/requirements.txt
train/model.py
train/model_new.py
train/train.py
train/train_new.py
train/requirements.txt
website/index.html