This is a subset (1k samples) of the SeyedAli/Persian-Text-QA dataset, processed to match Llama 2's prompt format as described in this article. It was created using the following colab notebook.
Useful if you don't want to reformat it by yourself (e.g., using a script). It was designed for this article about fine-tuning a Llama 2 (chat) model in a Google Colab.
