Spaces:
Sleeping
Sleeping
FrancoisHB
commited on
Commit
•
40be60a
1
Parent(s):
4d10fc5
Commit Test SRT
Browse files
app.py
CHANGED
@@ -1,7 +1,10 @@
|
|
1 |
import streamlit as st
|
|
|
2 |
from transformers import pipeline
|
3 |
from keybert import KeyBERT
|
4 |
|
|
|
|
|
5 |
|
6 |
# Function to extract text from SRT-formatted text
|
7 |
def extract_text_from_srt_text(srt_text):
|
@@ -9,7 +12,6 @@ def extract_text_from_srt_text(srt_text):
|
|
9 |
texts = [subtitle.split("\n")[2] for subtitle in lines if subtitle.strip()]
|
10 |
return " ".join(texts)
|
11 |
|
12 |
-
|
13 |
# Function to generate summary from text
|
14 |
def generate_summary(text, summary_length):
|
15 |
summarizer = pipeline("summarization")
|
@@ -17,43 +19,48 @@ def generate_summary(text, summary_length):
|
|
17 |
summary_text = summary[0]["summary_text"]
|
18 |
return summary_text
|
19 |
|
20 |
-
|
21 |
# Function to extract top 4 topics from text
|
22 |
def extract_top_topics(text, n_top_topics):
|
23 |
model = KeyBERT('distilbert-base-nli-mean-tokens')
|
24 |
-
keywords = model.extract_keywords(text, keyphrase_ngram_range=(1, 3), stop_words='english', use_maxsum=True,
|
25 |
-
|
26 |
-
return keywords
|
27 |
-
|
28 |
|
29 |
# Streamlit app
|
30 |
st.title("SRT Summarization")
|
31 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
32 |
|
33 |
-
# Define
|
34 |
-
@
|
35 |
-
def
|
|
|
|
|
|
|
|
|
36 |
text_to_summarize = extract_text_from_srt_text(srt_text)
|
37 |
summary = generate_summary(text_to_summarize, 150)
|
38 |
top_topics = extract_top_topics(text_to_summarize, 4)
|
39 |
-
return summary, top_topics
|
40 |
-
|
41 |
-
|
42 |
-
# Main entry point
|
43 |
-
def main():
|
44 |
-
srt_text_input = st.text_area("Paste SRT-formatted text here:")
|
45 |
-
|
46 |
-
if st.button("Summarize"):
|
47 |
-
if srt_text_input.strip():
|
48 |
-
summary, top_topics = summarize_text(srt_text_input)
|
49 |
-
st.subheader("Summary:")
|
50 |
-
st.write(summary)
|
51 |
-
st.subheader("Top 4 Keywords:")
|
52 |
-
for topic, _ in top_topics:
|
53 |
-
st.write(f"- {topic}")
|
54 |
-
else:
|
55 |
-
st.warning("Please enter some SRT-formatted text.")
|
56 |
-
|
57 |
|
58 |
if __name__ == "__main__":
|
59 |
-
|
|
|
1 |
import streamlit as st
|
2 |
+
from flask import Flask, request, jsonify
|
3 |
from transformers import pipeline
|
4 |
from keybert import KeyBERT
|
5 |
|
6 |
+
# Initialize Flask app
|
7 |
+
app = Flask(__name__)
|
8 |
|
9 |
# Function to extract text from SRT-formatted text
|
10 |
def extract_text_from_srt_text(srt_text):
|
|
|
12 |
texts = [subtitle.split("\n")[2] for subtitle in lines if subtitle.strip()]
|
13 |
return " ".join(texts)
|
14 |
|
|
|
15 |
# Function to generate summary from text
|
16 |
def generate_summary(text, summary_length):
|
17 |
summarizer = pipeline("summarization")
|
|
|
19 |
summary_text = summary[0]["summary_text"]
|
20 |
return summary_text
|
21 |
|
|
|
22 |
# Function to extract top 4 topics from text
|
23 |
def extract_top_topics(text, n_top_topics):
|
24 |
model = KeyBERT('distilbert-base-nli-mean-tokens')
|
25 |
+
keywords = model.extract_keywords(text, keyphrase_ngram_range=(1, 3), stop_words='english', use_maxsum=True, nr_candidates=20, top_n=n_top_topics)
|
26 |
+
return [topic for topic, _ in keywords]
|
|
|
|
|
27 |
|
28 |
# Streamlit app
|
29 |
st.title("SRT Summarization")
|
30 |
|
31 |
+
# Text area for user to input SRT-formatted text
|
32 |
+
srt_text_input = st.text_area("Paste SRT-formatted text here:")
|
33 |
+
|
34 |
+
# Button to trigger summarization
|
35 |
+
if st.button("Summarize"):
|
36 |
+
# Check if text area is not empty
|
37 |
+
if srt_text_input.strip():
|
38 |
+
# Extract text from SRT-formatted text
|
39 |
+
text_to_summarize = extract_text_from_srt_text(srt_text_input)
|
40 |
+
# Generate summary
|
41 |
+
summary = generate_summary(text_to_summarize, 150)
|
42 |
+
# Extract top 4 topics
|
43 |
+
top_topics = extract_top_topics(text_to_summarize, 4)
|
44 |
+
# Display summary and top 4 topics
|
45 |
+
st.subheader("Summary:")
|
46 |
+
st.write(summary)
|
47 |
+
st.subheader("Top 4 Keywords:")
|
48 |
+
for topic in top_topics:
|
49 |
+
st.write(f"- {topic}")
|
50 |
+
else:
|
51 |
+
st.warning("Please enter some SRT-formatted text.")
|
52 |
|
53 |
+
# Define endpoint for REST API
|
54 |
+
@app.route("/summarize", methods=["POST"])
|
55 |
+
def summarize():
|
56 |
+
data = request.json
|
57 |
+
if "srt_text" not in data:
|
58 |
+
return jsonify({"error": "Missing 'srt_text' parameter"}), 400
|
59 |
+
srt_text = data["srt_text"]
|
60 |
text_to_summarize = extract_text_from_srt_text(srt_text)
|
61 |
summary = generate_summary(text_to_summarize, 150)
|
62 |
top_topics = extract_top_topics(text_to_summarize, 4)
|
63 |
+
return jsonify({"summary": summary, "top_topics": top_topics})
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
64 |
|
65 |
if __name__ == "__main__":
|
66 |
+
app.run()
|