NLP with Transformers: Advanced Techniques and Multimodal ApplicationsChapter 43

Short Answer Questions

Section 3 of 4-~ 1 min read-Synced from Cuantum content
  1. Briefly explain why beam search is used when generating text summaries or translations.
  2. Provide an example of the Helsinki-NLP model name you would use for French-to-English translation.
  3. What is the purpose of the length_penalty parameter in text generation with T5?
  4. How would you handle text inputs that exceed the token limit of the T5 model?
  5. Explain one real-world use case of machine translation using MarianMT.