methods/searchain/baseline_eval.py

import os
from transformers import pipeline
import torch
import pandas as pd

from transformers import AutoTokenizer
import requests
from openai import OpenAI


dataset_path = os.path.join(os.path.dirname(os.path.abspath(__file__)), "data", "strategyqa_dev.csv")

raw_data = pd.read_csv(dataset_path)
raw_queries = list(raw_data['question'])
true_answers = list(raw_data['answer'])
true_answers = ['yes' if row else 'no' for row in true_answers]

acc_dic = {}

def send_request(FLASK_URL, questions, answers):
    
    payload = {
        "questions": questions,
        "answers": answers
    }
    url = "http://127.0.0.1:50008/execute"

  
    params = {
        "query": questions, 
        "answers": answers
    }

    # 发送请求
    response = requests.get(url, params=params, timeout=None)
    
    response_data = response.json()

    query_ls = response_data.get("query_ls", [])
    ans_ls = response_data.get("ans_ls", [])
    acc_ls = response_data.get("acc_ls", [])
    return query_ls, ans_ls, acc_ls


def get_acc(queries, answers):

    FLASK_URL = "http://127.0.0.1:50008/execute"
    query_ls, ans_ls, acc_ls = send_request(FLASK_URL, raw_queries, true_answers)
    
    average = sum(acc_ls) / len(acc_ls)
    return average




acc = get_acc( raw_queries, true_answers)
print(f"baseline, accuracy: {acc}")