How to beginner · 4 min read

How to use summarization pipeline Hugging Face

Q: How to use summarization pipeline Hugging Face

Use the Hugging Face transformers library's pipeline function with the task set to summarization. Load a pretrained model like facebook/bart-large-cnn and pass your text to get a concise summary.

Quick answer

Use the Hugging Face transformers library's pipeline function with the task set to summarization. Load a pretrained model like facebook/bart-large-cnn and pass your text to get a concise summary.

PREREQUISITES

Python 3.8+
pip install transformers>=4.0.0
pip install torch (or tensorflow)
Internet connection to download pretrained models

Setup

Install the transformers library and a backend like torch or tensorflow. Set up your Python environment to run the summarization pipeline.

bash

pip install transformers torch

Step by step

Use the pipeline API from transformers to create a summarization pipeline with a pretrained model. Pass your input text to get a summary.

python

from transformers import pipeline

# Initialize summarization pipeline with a pretrained model
summarizer = pipeline('summarization', model='facebook/bart-large-cnn')

# Input text to summarize
text = ("The Hugging Face Transformers library provides thousands of pretrained models to perform tasks on texts such as classification, information extraction, question answering, summarization, translation, and more.")

# Generate summary
summary = summarizer(text, max_length=50, min_length=25, do_sample=False)

print('Summary:', summary[0]['summary_text'])

output

Summary: The Hugging Face Transformers library provides pretrained models to perform tasks on texts such as classification, information extraction, question answering, summarization, translation, and more.

Common variations

Use different models like t5-small or google/pegasus-xsum for summarization.
Adjust max_length and min_length parameters to control summary length.
Use the pipeline asynchronously with asyncio for batch processing.

python

from transformers import pipeline
import asyncio

async def async_summarize(text):
    summarizer = pipeline('summarization', model='facebook/bart-large-cnn')
    # Run summarization asynchronously
    summary = await asyncio.to_thread(summarizer, text, max_length=60, min_length=30, do_sample=False)
    print('Async summary:', summary[0]['summary_text'])

text = "Hugging Face provides easy-to-use APIs for state-of-the-art NLP models."
asyncio.run(async_summarize(text))

output

Async summary: Hugging Face provides easy-to-use APIs for state-of-the-art NLP models.

Troubleshooting

If you see Model not found, verify the model name and internet connection.
For CUDA out of memory errors, reduce batch size or run on CPU by setting device=-1 in the pipeline.
If summaries are too short or too long, adjust min_length and max_length parameters.

✅

Key Takeaways

Use Hugging Face's pipeline with task summarization for easy text summarization.
Choose pretrained models like facebook/bart-large-cnn for high-quality summaries.
Adjust max_length and min_length to control summary size.
Handle errors by checking model names, internet connection, and device memory.
Async summarization can be done using Python's asyncio with pipeline.

Verified 2026-04 · facebook/bart-large-cnn, t5-small, google/pegasus-xsum

Verify ↗