Building a Q&A pipeline
In the previous exercise, we extracted text from a multi-page document using PyPDF and prepared it as a single string.
Now, you’ll build a question-answering pipeline using Hugging Face to retrieve specific answers from the document.
Both the pipeline module and document_text have been preloaded for you.
This exercise is part of the course
Working with Hugging Face
Exercise instructions
- Initialize a question-answering pipeline with the correct task and model.
- Pass the question and
document_textto the pipeline. - Print the result to view the answer.
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# Load the question-answering pipeline
qa_pipeline = pipeline(task="____", model="distilbert-base-cased-distilled-squad")
question = "What is the notice period for resignation?"
# Get the answer from the QA pipeline
result = qa_pipeline(question=____, context=____)
# Print the answer
print(f"Answer: {result['____']}")