File size: 4,734 Bytes
1550086 |
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 |
import os
import streamlit as st
import faiss
import pandas as pd
import numpy as np
import requests
import torch
import re
import time
from sentence_transformers import SentenceTransformer, CrossEncoder
from huggingface_hub import hf_hub_download
from langdetect import detect
HF_REPO = "AKKI-AFK/deepshelf-data"
books_file = hf_hub_download(repo_id=HF_REPO, filename="booksummaries.txt", repo_type="dataset")
faiss_file = hf_hub_download(repo_id=HF_REPO, filename="faiss_index.bin", repo_type="dataset")
df = pd.read_csv(books_file, delimiter="\t")
index = faiss.read_index(faiss_file)
encoder = SentenceTransformer("sentence-transformers/paraphrase-mpnet-base-v2")
cross_encoder = CrossEncoder("cross-encoder/ms-marco-MiniLM-L6-v2")
request_times = [] # Track request timestamps
@st.cache_data
def recommend_books(query):
query = sanitize_input(query)
if len(query) > 200:
st.warning("⚠️ Query is too long. Please keep it under 200 characters.")
return []
if len(query) < 3:
st.warning("⚠️ Query is too short. Please provide more details.")
return []
try:
lang = detect(query)
if lang != "en":
st.warning("⚠️ Non-English query detected. Results may not be accurate.")
except:
st.warning("⚠️ Could not detect language. Ensure proper input.")
search_vector = encoder.encode(query)
search_vector = np.array([search_vector])
faiss.normalize_L2(search_vector)
distances, ann = index.search(search_vector, k=50)
results = pd.DataFrame({'distances': distances[0], 'ann': ann[0]})
merge = pd.merge(results, df, left_on='ann', right_index=True)
merge['Query'] = query
pairs = list(zip(merge['Query'], merge['summary']))
scores = cross_encoder.predict(pairs) # Running synchronously
merge['score'] = scores
df_sorted = merge.iloc[merge["score"].argsort()][::-1]
return df_sorted[["title", "summary"]][:5].to_dict(orient="records")
def sanitize_input(text):
"""Sanitize input by removing special characters and excessive spaces."""
text = re.sub(r'[^\w\s]', '', text) # Remove special characters
text = re.sub(r'\s+', ' ', text).strip() # Normalize spaces
return text
def rate_limit():
"""Rate-limiting function to prevent excessive queries."""
global request_times
current_time = time.time()
request_times = [t for t in request_times if current_time - t < 10] # Keep only recent requests within 10 seconds
if len(request_times) >= 5:
st.error("⚠️ Too many requests. Please wait a few seconds before trying again.")
return False
request_times.append(current_time)
return True
st.set_page_config(page_title="DeepShelf", page_icon="📚", layout="wide")
st.markdown("""
<style>
body {background-color: #1E1E1E; color: white;}
.title {text-align: center; font-size: 3em; font-weight: bold; color: #E6A400;}
.subtext {text-align: center; font-size: 1.2em; color: #AAAAAA;}
.recommend-btn {text-align: center;}
.book-container {border-radius: 10px; padding: 20px; margin: 10px; background: #2E2E2E; box-shadow: 2px 2px 10px #00000050;}
.book-title {font-size: 1.5em; font-weight: bold; color: #FFD700;}
.book-summary {font-size: 1em; color: #CCCCCC;}
</style>
""", unsafe_allow_html=True)
st.markdown('<div class="title">📖 DeepShelf</div>', unsafe_allow_html=True)
st.markdown('<div class="subtext">Find the best books based on your preferences!</div>', unsafe_allow_html=True)
query = st.text_input("🔍 Enter a book description (e.g., 'A dark fantasy with drama')", max_chars=200, help="Use keywords to describe your ideal book!")
button_disabled = not (3 <= len(query) <= 200) # Disable button if query length is invalid
button_placeholder = st.empty()
if button_placeholder.button("✨ Recommend Books", disabled=button_disabled, help="Click to get personalized book recommendations!"):
if rate_limit():
button_placeholder.button("✨ Processing...", disabled=True) # Disable button while processing
with st.spinner("🔍 Searching for the best books..."):
recommendations = recommend_books(query)
button_placeholder.button("✨ Recommend Books", disabled=False) # Re-enable button after processing
st.markdown("## 📚 Recommended Books:")
for rec in recommendations:
st.markdown(f"""
<div class="book-container">
<div class="book-title">📖 {rec["title"]}</div>
<div class="book-summary">{rec["summary"]}</div>
</div>
""", unsafe_allow_html=True) |