Datasets:
extracted and reformatted for LLama 2 from SALT-NLP/FLUE-FiQA for easier use.
Some problems that might exist in the dataset:
- unnecessary information and context that can affect training
- emotions/non-objective tone that will be inherited by model in training
- information inside might not be up-to-date
- Downloads last month
- 31