Create summarize.py
Browse files- summarize.py +29 -0
summarize.py
ADDED
|
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
"""Build a Hugging Face Summarize Library"""
|
| 2 |
+
|
| 3 |
+
from transformers import pipeline
|
| 4 |
+
import wikipedia
|
| 5 |
+
|
| 6 |
+
|
| 7 |
+
def summarize(text):
|
| 8 |
+
"""Summarize text"""
|
| 9 |
+
summarizer = pipeline("summarization", model="t5-small")
|
| 10 |
+
result = summarizer(text, max_length=180)
|
| 11 |
+
print("Summarization process complete!")
|
| 12 |
+
print("=" * 80)
|
| 13 |
+
return result[0]["summary_text"]
|
| 14 |
+
|
| 15 |
+
|
| 16 |
+
# search wikipedia pages
|
| 17 |
+
def search_wiki(search_term):
|
| 18 |
+
"""Search wikipedia pages"""
|
| 19 |
+
|
| 20 |
+
search_results = wikipedia.search(search_term)
|
| 21 |
+
return search_results
|
| 22 |
+
|
| 23 |
+
|
| 24 |
+
# grab text from a wikipedia page
|
| 25 |
+
def get_wiki_text(page_title):
|
| 26 |
+
"""Grab text from a wikipedia page"""
|
| 27 |
+
|
| 28 |
+
page = wikipedia.page(page_title)
|
| 29 |
+
return page.content
|