Description

Natural Language Processing)
25% of the course mark
23:59
1. Develop a sentiment classifier with 3 classes (pro-vax, anti-vax, neutral) by fine-tuning the pretrained BERT-base model available on Hugging Face. . You have to compute precision, recall and F1 for each class.
(3/10 marks)
(4/10 marks)
3. In this question you will read the paper “What do Models Learn from Question Answering Datasets?”3 and you will reproduce some of the results of that paper. In particular, you have to reproduce the results of Table 3 (at least the rows and columns for SQuAD and TriviaQA).
(3-6/10 marks depending on how many datasets from Table 3 you will cover)
Note: Your solutions should be implemented in PyTorch. You should hand in: (i) a pdf with a detailed report of your solution for the two questions, including an explanation of the methods you used, an evaluation and comparison of different approaches you tried and citations to relevant literature that you might have used in developing your solutions. (ii)
Colab notebooks (ipynb files using https://colab.research.google.com/) containing your code. You should use Python 3.6 or a later version.
1