SKILL BENCHMARK

NLP and LLMs Competency (Intermediate Level)

  • 19m
  • 19 questions
Explore all Skills Benchmarks
The NLP and LLMs Competency (Intermediate Level) benchmark measures your knowledge of working with tokenizers in Hugging Face. You will be evaluated on your recognition of Hugging Face classification, QnA pipelines, and text generation pipelines. A learner who scores high on this benchmark demonstrates that they have good experience in developing NLP and LLM applications using Hugging Face with minimal supervision.

Topics covered

  • explore greedy search and beam search for text generation
  • explore normalization and pre-tokenization
  • get predictions from a QnA pipeline
  • implement BPE tokenization
  • implement WordPiece tokenization
  • outline the use of Hugging Face pipelines
  • perform byte pair encoding (BPE) and WordPiece tokenization
  • perform mask filling using a BERT model
  • perform named entity recognition (NER) with a BERT model
  • perform NER manually
  • perform normalization and pre-tokenization with WordPiece
  • perform sentiment analysis with FinBERT
  • perform text generation with sampling
  • perform zero-shot classification
  • set up a BPE tokenizer
  • set up a WordPiece tokenizer
  • train a BPE tokenizer
  • train a WordPiece tokenizer
  • use a QnA pipeline