# Summarization pipeline

**URL:** <https://discuss.huggingface.co/t/summarization-pipeline/58897>\
**Category:** 🤗Transformers\
**Created:** [October 17, 2023, 3:58pm UTC](https://discuss.huggingface.co/t/summarization-pipeline/58897 "2023-10-17T15:58:11Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![savasci](https://avatars.discourse-cdn.com/v4/letter/s/e9c0ed/32.png) [@savasci](https://discuss.huggingface.co/u/savasci)\
**Post date:** [October 17, 2023, 3:58pm UTC](https://discuss.huggingface.co/t/summarization-pipeline/58897/1 "2023-10-17T15:58:11Z")

</div>

Hi everyone,

I’m testing the summarization pipeline that is explained [here](https://huggingface.co/learn/nlp-course/chapter1/3?fw=pt)  
I want a summarization model that extracts key phrases from the text. On the contrary, the generated summaries using this pipeline include sentences that are not in the text (in other words, it generates a text\_ or summary\_ that in the meaning is close to the original one).  
Is there any way to change its approach?

Thanks