Explain Transformer, GPT vs BERT, and PR metrics

Respond to the conceptual questions below:

1. Transformer architecture
   • Outline the principal parts of a Transformer block and explain the role of each one.
2. GPT vs BERT
   • Describe the main differences in how the architectures are used and in their pretraining goals.
   • In what situations would you choose one instead of the other?
3. Precision and recall
   • Give the definitions of precision and recall.
   • Provide an example showing how adjusting a threshold can trade precision against recall.
   • Name at least one situation where precision matters more and one where recall matters more.

Overview: This question assesses familiarity with modern NLP model architectures and evaluation metrics: specifically, the components of a Transformer block, the main differences between GPT and BERT including their pretraining objectives and usage, and the interpretation of precision and recall. It tests deep learning architecture knowledge, model selection reasoning, and metric-based evaluation within the Software Engineering Fundamentals domain for a Machine Learning Engineer role. The question is common because it examines both conceptual knowledge of architectures and pretraining paradigms and the practical handling of performance trade-offs, evaluating conceptual understanding together with practical use of threshold-based precision-versus-recall choices.