metadata
task_categories:
- question-answering
language:
- en
tags:
- finance
pretty_name: tiny_FAQs
size_categories:
- n<1K
This is the dataset used to fine tuning (train) the LLama2. This dataset has the original questios (FAQs), with only 37 samples and 2 features (questions and answers). This version dataset is in ENLIGSH:
Another dataset was created to increase the number of samples by applying data augmentation (synthetic data). Check dataset in Fraternitas repo