https://colab.research.google.com/drive/1gvGBM12RdHQoJG48bguIs2h2MoQ6sxJB?usp=sh

https://colab.research.google.com/drive/1gvGBM12RdHQoJG48bguIs2h2MoQ6sxJB?usp=sharing As indicated on the notebook: “In this assignment you have to build and train a multimodal deep neural network for sentiment recognition using tf.keras/pytorch. You have to work with the MOSI dataset, which contains more than 2000 short videos.”
I need the following task to be done: “classification with 2 classes: [-3, +3] values are converted to 2 classes: negative and positive.”. No need to do the optional tasks but please note that the base task that must be done requires to build a MULTIMODAL deep neural network for the classification..