Json – How should I preprocess this dataset for performing a "Question-Answering" task? Pytorch
So I have a JSON dataset that is in the following format: [{ "answer": "...", "question": "...", "context": "..." }, ... ] All of the fields are normal plaintext. My goal is train a pretrained BERT model on it for…