All the submissions are considered and ranked based on macro average F1 score, determining their performance across multiple classes. The submissions are then ordered based on this metric to identify the best-performing models.
To view the results: click here
To view the run-wise results: click here