initial version

app.py

@@ -0,0 +1,44 @@
+# import
+from tensorflow.python.keras.utils.generic_utils import default
+import streamlit as st
+from newspaper import Article
+from transformers import pipeline
+
+st.set_page_config(layout="wide", page_title="SummarizeLink")
+
+# load the summarization model
+@st.cache(allow_output_mutation=True)
+def load_summarize_model():
+    # model = pipeline("summarization", model='sshleifer/distilbart-cnn-12-6')
+    model = pipeline("summarization")
+    return model
+summ = load_summarize_model()
+
+# define functions
+def download_and_parse_article(url):
+    article = Article(url)
+    article.download()
+    article.parse()
+    return article.text
+
+# define the app
+st.title("SummarizeLink")
+st.text("Paste any article link below and click on the 'Summarize Text' button to get the summarized data")
+# st.subheader("This application is using HuggingFace's transformers pre-trained model for text summarization.")
+link = st.text_area('Paste your link here...', "https://towardsdatascience.com/a-guide-to-the-knowledge-graphs-bfb5c40272f1", height=50)
+button = st.button("Summarize")
+max_lengthy = st.sidebar.slider('Max summary length', min_value=30, max_value=700, value=100, step=10)
+# num_beamer = st.sidebar.slider('Speed vs quality of Summary (1 is fastest but less accurate)', min_value=1, max_value=8, value=4, step=1)
+with st.spinner("Summarizing..."):
+    if button and link:
+        text = download_and_parse_article(link)     # get the text
+        summary = summ(text, 
+                       truncation=True,
+                       max_length = max_lengthy, 
+                       min_length = 50, 
+                       num_beams=5, 
+                       do_sample=True,
+                       early_stopping=True, 
+                       repetition_penalty=1.5, 
+                       length_penalty=1.5)[0]
+        st.write(summary['summary_text'])

requirements.txt

@@ -0,0 +1,2 @@
+streamlit
+newspaper