app.py

import gradio as gr
import pandas as pd
from sentence_transformers import SentenceTransformer
from sklearn.metrics.pairwise import cosine_similarity

title = "🍀고민 해결 도서 추천 챗봇🍀"
description = "고민이 무엇인가요? 고민 해결을 도와줄 책을 추천해드립니다"
examples = [["요즘 잠이 안 온다"]]


model = SentenceTransformer('jhgan/ko-sroberta-multitask')

def response(message):
  embedding = model.encode(message)
  df['distance'] = df['embedding'].map(lambda x: cosine_similarity([embedding], [x]).squeeze())
  answer = df.loc[df['distance'].idxmax()]
  Book_title = answer['제목']
  Book_author = answer['작가']
  Book_publisher = answer['출판사']
  Book_comment = answer['서평']
  return print(message)


gr.Interface(
    fn=predict,
    title=title,
    description=description,
    examples=examples,
    inputs=["text", "state"],
    outputs=["chatbot", "state"],
    theme="finlaymacklon/boxy_violet",
).launch()