Skip to content

Latest commit

 

History

History
51 lines (34 loc) · 1.6 KB

README.md

File metadata and controls

51 lines (34 loc) · 1.6 KB

Multi-Label Movie Genre Classification

Predict genre names by looking at the summary of the movie.

Usage:

Step 1:

docker-compose build
docker-compose up

WARNING: Before executing the above commands, make sure that in plot_summary_genre_api directory, there is a model inside models directory. For example:

plot_summary_genre_api
|-- README.md
`-- models
   `-- bert  	<---------------This is the model name
       `-- 2 	<---------------This is the version number
           |-- assets
           |-- saved_model.pb
           `-- variables
               |-- variables.data-00000-of-00001
               `-- variables.index

Download the fine-tuned bert model.

Step 2:

Go to the webpage at http://localhost:8080/

Sample Movie Plot:

This movie is based on a story of two lovers who meet on a segway. However, the world is not happy with them being together. So, the world keeps on making strategies to keep them from coming together. Even after these obstacles, they manage to elope and live a happy life together.

Services:

There are 3 services: web, app, plot_summary_genre_api

web:

This service acts as a website server and displays a User Interface to let the user interact with the app and predict movie genres.

app:

This is where the pre-processing of the user request and the post-processing of the model api response happens.

plot_summary_genre_api:

This is where the model is hosted as an API using TF Serving