File size: 4,277 Bytes
716037e 71754ec 716037e 71754ec 716037e 71754ec 716037e 71754ec 716037e 71754ec 716037e 71754ec 716037e 71754ec 716037e 71754ec 716037e 71754ec 716037e 71754ec 716037e 71754ec 716037e 71754ec 716037e 71754ec |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 |
import gradio as gr
import logging
import sys
import os
import gc
# Configure logging
logging.basicConfig(
level=logging.INFO,
format='%(asctime)s - %(name)s - %(levelname)s - %(message)s',
handlers=[logging.StreamHandler(sys.stdout)]
)
logger = logging.getLogger(__name__)
# Log startup information
logger.info("Starting StudAI Summarization Service with Gradio")
logger.info(f"Python version: {sys.version}")
# Force garbage collection
gc.collect()
# Create a simple function for summarization that doesn't use ML in case model loading fails
def simple_summarize(text, max_length=150, min_length=30):
"""Simple extractive summarization as fallback"""
import re
sentences = re.split(r'(?<=[.!?])\s+', text)
if len(sentences) <= 3:
return text
# Take first, middle and last sentences
summary = [
sentences[0],
sentences[len(sentences) // 2],
sentences[-1]
]
return " ".join(summary)
# Set a flag for model availability
model_available = False
# Try to import and load the model with memory optimizations
try:
# Import and load only when needed
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM, pipeline
import torch
logger.info("Loading small model for summarization...")
# Use a tiny model instead of t5-small
model_name = "facebook/bart-large-cnn"
# Enable memory optimization
if torch.cuda.is_available():
logger.info("CUDA available, using GPU")
device = 0
else:
logger.info("CUDA not available, using CPU")
device = -1
# Enable memory-efficient loading
summarizer = pipeline(
"summarization",
model=model_name,
device=device,
framework="pt"
)
logger.info("Model loaded successfully!")
model_available = True
# Force garbage collection after model loading
gc.collect()
if torch.cuda.is_available():
torch.cuda.empty_cache()
except Exception as e:
logger.error(f"Failed to load model: {str(e)}")
logger.info("Will use simple extractive summarization instead")
def summarize_text(text, max_length=150, min_length=30):
"""Summarize the provided text"""
if not text or len(text.strip()) < 50:
return text
try:
if model_available:
logger.info(f"Summarizing text of length {len(text)} with model")
result = summarizer(
text,
max_length=max_length,
min_length=min_length,
truncation=True
)
summary = result[0]["summary_text"]
else:
logger.info(f"Using simple summarization for text of length {len(text)}")
summary = simple_summarize(text, max_length, min_length)
return summary
except Exception as e:
logger.error(f"Error during summarization: {str(e)}")
# Fall back to simple summarization on error
return simple_summarize(text, max_length, min_length)
# Create Gradio interface
demo = gr.Interface(
fn=summarize_text,
inputs=[
gr.Textbox(
lines=10,
label="Text to Summarize",
placeholder="Enter text to summarize (at least 50 characters)"
),
gr.Slider(50, 500, value=150, label="Max Length"),
gr.Slider(10, 200, value=30, label="Min Length")
],
outputs=gr.Textbox(label="Summary"),
title="StudAI Text Summarization",
description="This service provides text summarization for the StudAI Android app.",
examples=[
["The coronavirus pandemic has led to a surge in remote work. Companies around the world have had to adapt to new ways of working, with many employees setting up home offices. This shift has led to changes in productivity, work-life balance, and communication patterns. Some studies suggest that remote work can increase productivity, while others point to challenges in collaboration and team cohesion. Organizations are now considering hybrid models for the future of work.", 150, 30]
],
allow_flagging="never"
)
# Launch with parameters optimized for Spaces
demo.launch(share=False, server_name="0.0.0.0", server_port=7860) |