sugiv/

Unified_Multi-task_NLP_Dataset

This is a multi-task NLP dataset designed for training and evaluating language models across different domains including QA evaluation, command generation, and entity extraction.

This dataset combines multiple sources:

CALM Command Generation Dataset Synthetic QA Evaluation Dataset (full, partial, none) Synthetic Entity Extraction Dataset

All examples have been validated using the DeepSeek API to ensure coherence and quality.… See the full description on the dataset page: https://huggingface.co/datasets/sugiv/Unified_Multi-task_NLP_Dataset.