Update app.py
Browse files
app.py
CHANGED
|
@@ -7,7 +7,7 @@ import requests
|
|
| 7 |
from bs4 import BeautifulSoup
|
| 8 |
import concurrent.futures
|
| 9 |
import time
|
| 10 |
-
import
|
| 11 |
import io
|
| 12 |
import base64
|
| 13 |
|
|
@@ -17,9 +17,6 @@ logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(
|
|
| 17 |
summarizer = pipeline("summarization", model="t5-small", tokenizer="t5-small")
|
| 18 |
tokenizer = AutoTokenizer.from_pretrained("t5-small")
|
| 19 |
|
| 20 |
-
# Initialize pyttsx3
|
| 21 |
-
engine = pyttsx3.init()
|
| 22 |
-
|
| 23 |
def fetch_content_from_url(url):
|
| 24 |
try:
|
| 25 |
response = requests.get(url, timeout=10)
|
|
@@ -35,93 +32,4 @@ def chunk_text(text, max_chunk_size=200):
|
|
| 35 |
words = text.split()[:1000] # Limit to first 1000 words
|
| 36 |
chunks = []
|
| 37 |
current_chunk = []
|
| 38 |
-
current_size
|
| 39 |
-
for word in words:
|
| 40 |
-
if current_size + len(word) > max_chunk_size:
|
| 41 |
-
chunks.append(' '.join(current_chunk))
|
| 42 |
-
current_chunk = [word]
|
| 43 |
-
current_size = len(word)
|
| 44 |
-
else:
|
| 45 |
-
current_chunk.append(word)
|
| 46 |
-
current_size += len(word) + 1
|
| 47 |
-
if current_chunk:
|
| 48 |
-
chunks.append(' '.join(current_chunk))
|
| 49 |
-
return chunks
|
| 50 |
-
|
| 51 |
-
def summarize_chunk(chunk):
|
| 52 |
-
try:
|
| 53 |
-
summary = summarizer(chunk, max_length=50, min_length=10, do_sample=False)[0]['summary_text']
|
| 54 |
-
return summary
|
| 55 |
-
except Exception as e:
|
| 56 |
-
logging.error(f"Error summarizing chunk: {str(e)}")
|
| 57 |
-
return ""
|
| 58 |
-
|
| 59 |
-
def text_to_speech(text):
|
| 60 |
-
try:
|
| 61 |
-
engine.save_to_file(text, 'summary.mp3')
|
| 62 |
-
engine.runAndWait()
|
| 63 |
-
with open("summary.mp3", "rb") as audio_file:
|
| 64 |
-
audio_base64 = base64.b64encode(audio_file.read()).decode('utf-8')
|
| 65 |
-
return f'data:audio/mp3;base64,{audio_base64}'
|
| 66 |
-
except Exception as e:
|
| 67 |
-
logging.error(f"Error in text-to-speech conversion: {str(e)}")
|
| 68 |
-
return None
|
| 69 |
-
|
| 70 |
-
def summarize_text(input_text, is_url):
|
| 71 |
-
try:
|
| 72 |
-
start_time = time.time()
|
| 73 |
-
if is_url:
|
| 74 |
-
content = fetch_content_from_url(input_text)
|
| 75 |
-
if content.startswith("Error"):
|
| 76 |
-
return content, None
|
| 77 |
-
else:
|
| 78 |
-
content = input_text[:10000] # Limit direct input to 10000 characters
|
| 79 |
-
|
| 80 |
-
chunks = chunk_text(content)
|
| 81 |
-
with concurrent.futures.ThreadPoolExecutor() as executor:
|
| 82 |
-
summaries = list(executor.map(summarize_chunk, chunks))
|
| 83 |
-
|
| 84 |
-
full_summary = " ".join(summaries)
|
| 85 |
-
|
| 86 |
-
processing_time = time.time() - start_time
|
| 87 |
-
summary_with_time = f"Summary (processed in {processing_time:.2f} seconds):\n\n{full_summary}"
|
| 88 |
-
|
| 89 |
-
audio = text_to_speech(full_summary)
|
| 90 |
-
|
| 91 |
-
return summary_with_time, audio
|
| 92 |
-
except Exception as e:
|
| 93 |
-
error_msg = f"An error occurred: {str(e)}\n\nTraceback:\n{traceback.format_exc()}"
|
| 94 |
-
logging.error(error_msg)
|
| 95 |
-
return error_msg, None
|
| 96 |
-
|
| 97 |
-
def debug_info():
|
| 98 |
-
return (
|
| 99 |
-
f"Python version: {sys.version}\n"
|
| 100 |
-
f"Summarizer model: T5-small\n"
|
| 101 |
-
f"TTS Engine: pyttsx3\n"
|
| 102 |
-
f"Current working directory: {os.getcwd()}\n"
|
| 103 |
-
f"Contents of current directory: {os.listdir()}"
|
| 104 |
-
)
|
| 105 |
-
|
| 106 |
-
iface = gr.Interface(
|
| 107 |
-
fn=summarize_text,
|
| 108 |
-
inputs=[
|
| 109 |
-
gr.Textbox(lines=5, label="Input Text or URL"),
|
| 110 |
-
gr.Checkbox(label="Is URL?")
|
| 111 |
-
],
|
| 112 |
-
outputs=[
|
| 113 |
-
gr.Textbox(label="Summary", lines=10),
|
| 114 |
-
gr.Audio(label="Audio Summary")
|
| 115 |
-
],
|
| 116 |
-
title="Quick Text Summarizer with Speech",
|
| 117 |
-
description="Enter text or a URL to summarize. Long texts will be truncated for quick processing."
|
| 118 |
-
)
|
| 119 |
-
|
| 120 |
-
debug_interface = gr.Interface(
|
| 121 |
-
fn=debug_info,
|
| 122 |
-
inputs=None,
|
| 123 |
-
outputs="text",
|
| 124 |
-
title="Debug Information"
|
| 125 |
-
)
|
| 126 |
-
|
| 127 |
-
gr.TabbedInterface([iface, debug_interface], ["Summarizer", "Debug Info"]).launch()
|
|
|
|
| 7 |
from bs4 import BeautifulSoup
|
| 8 |
import concurrent.futures
|
| 9 |
import time
|
| 10 |
+
from gtts import gTTS
|
| 11 |
import io
|
| 12 |
import base64
|
| 13 |
|
|
|
|
| 17 |
summarizer = pipeline("summarization", model="t5-small", tokenizer="t5-small")
|
| 18 |
tokenizer = AutoTokenizer.from_pretrained("t5-small")
|
| 19 |
|
|
|
|
|
|
|
|
|
|
| 20 |
def fetch_content_from_url(url):
|
| 21 |
try:
|
| 22 |
response = requests.get(url, timeout=10)
|
|
|
|
| 32 |
words = text.split()[:1000] # Limit to first 1000 words
|
| 33 |
chunks = []
|
| 34 |
current_chunk = []
|
| 35 |
+
current_size
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|