You will work with three subsets of the dataset: **train_dataset**, **val_dataset**, and **test_dataset**, which are used for training, validation, and testing, respectively.  Each one consists of three columns:

- **ID** : A unique identifier for each tweet.
- **Text** : The content of the tweet.
- **Label** : A binary value where **0** represents a **negative** sentiment and **1**  represents a **positive** sentiment.

Additionally, there is a file called **sample_submission.csv**, which serves as a template for submissions. It provides an example of the required format for the submission CSV file.