Seq2seq transformer from scratch for generative question answering (QA) task.
Using Science Questions Dataset (SciQ) from Hugging Face.
(Link: https://huggingface.co/datasets/allenai/sciq)
SciQ dataset is a multiple-choice question answering dataset focused on science topics typically taught at the middle school level. Contains over 13,000 crowd-sourced questions, each with a correct answer and three distractors (incorrect options). Covering diverse topics such as Physics, Chemistry, Biology, and among others. For the majority of the questions, an additional paragraph with supporting evidence for the correct answer is provided.
Preview dataset:
