CNN model laughs at a meme

Inspiration

In this experiment, a machine learning algorithm tries to predict ”how funny” a meme is. By training the algorithm on a dataset of memes from Reddit, the machine learns the popularity of a meme based on its number of upvotes. It is then able to predict how funny an original meme is through a Web user interface.

How we built it

The model is a convolutional neural network built with tensorflow on Python. The dataset was taken from kaggle at : https://www.kaggle.com/sayangoswami/reddit-memes-dataset

Built With

• cnn
• machine-learning
• python
• tensorflow