Consolidating and Developing Benchmarking Datasets for the Nepali Natural Language Understanding Tasks