Spaces:
Sleeping
Sleeping
| import streamlit as st | |
| from streamlit import session_state | |
| import os | |
| import openai | |
| openai.api_key = os.getenv("OPENAI_API_KEY") | |
| import pandas as pd | |
| from sklearn.preprocessing import LabelEncoder | |
| import numpy as np | |
| def gpt4_score(m_answer, s_answer): | |
| response = openai.ChatCompletion.create( | |
| model="gpt-4", | |
| messages=[ | |
| { | |
| "role": "system", | |
| "content": "You are UPSC answers evaluater. You will be given model answer and student answer. Evaluate it by comparing with the model answer. \n<<REMEMBER>>\nIt is 10 marks question. Student can recieve maximum 5 marks. Give marks in the range of 0.25. (ex. 0,0.25,0.5...)\nThere are 3 parts in the answer. Introduction (1 marks), body (3 marks) and conclusion (1 marks). If the student answer and model answer is not relevant then give 0 marks.\ngive output in json form. Give output in this format {\"intro\":,\"body\":,\"con\":,\"total\":}\n<<OUTPUT>>\n" | |
| }, | |
| { | |
| "role": "assistant", | |
| "content": f"Model answer: {m_answer}"}, | |
| { | |
| "role": "user", | |
| "content": f"Student answer: {s_answer}"} | |
| ], | |
| temperature=0, | |
| max_tokens=701, | |
| top_p=1, | |
| frequency_penalty=0, | |
| presence_penalty=0 | |
| ) | |
| return response.choices[0].message.content | |
| st.write("# Auto Score Generation! 👋") | |
| if 'score' not in session_state: | |
| session_state['score']= "" | |
| text1= st.text_area(label= "Please write the Model Answer bellow", | |
| placeholder="What does the text say?") | |
| text2= st.text_area(label= "Please write the Student Answer bellow", | |
| placeholder="What does the text say?") | |
| def classify(text1,text2): | |
| session_state['score'] = gpt4_score(text1,text2) | |
| st.text_area("result", value=session_state['score']) | |
| st.button("Submit", on_click=classify, args=[text1,text2]) |