basic_utils.py

import os
import re
import json
import numpy as np
import pandas as pd
from collections import Counter
from typing import Any, Optional, Tuple, Dict, List, NamedTuple, Set
import scipy
import time
from pprint import pprint as pprint
import openai  # Import the openai library directly

openai.api_key = os.getenv("OPENAI_API_KEY")
# Load your API key from an environment variable or secret management service
model_to_use = "gpt-3.5-turbo-0125"


# "gpt-3.5-turbo-0301" is more moderns
# "text-curie-001" is 10x cheaper than "text-davinci-003", but not as good
# "text-davinci-002" is faster than 3, but not necessarily worse at explaining?

# Setup interface with language model

# def gen_response(prompt='hi', max_tokens=10, temperature=0):
#     response = client.chat.completions.create(model=model_to_use, prompt=prompt, temperature=temperature, max_tokens=max_tokens)
#     return response

def gen_response(prompt='hi', max_tokens=10, temperature=0):
    response = openai.chat.completions.create(
        model=model_to_use,
        messages=[{"role": "user", "content": prompt}],
        temperature=temperature,
        max_tokens=max_tokens
    )
    # print(response)
    return response


def gen_response_text_with_backoff(prompt='hi', max_tokens=10, temperature=0):
    prompt_succeeded = False
    wait_time = 0.1
    while not prompt_succeeded:
        try:
            response = gen_response(prompt, max_tokens, temperature=temperature)
            prompt_succeeded = True
        except openai.error.RateLimitError as e:
            print('Rate limit exceeded, retrying after', wait_time, 'seconds...')
            time.sleep(wait_time)
            wait_time *= 2  # Exponential backoff
        except Exception as e:
            print('Unexpected error:', str(e), '— Retrying after', wait_time, 'seconds...')
            time.sleep(wait_time)
            wait_time *= 2  # Exponential backoff
    response_text = response.choices[0].message.content.strip()
    used_tokens = response.usage.total_tokens
    return response_text, used_tokens


def get_dict_items_sorted_by_decreasing_value(_dict):
    sort_inds = np.flip(np.argsort(list(_dict.values())))
    sorted_keys = np.array(list(_dict.keys()))[sort_inds]
    sorted_values = np.array(list(_dict.values()))[sort_inds]
    return sorted_keys, sorted_values


def display_dict_sorted_by_decreasing_value(_dict, print_num=10):
    sorted_keys, sorted_values = get_dict_items_sorted_by_decreasing_value(_dict)
    pprint(list(zip(sorted_keys, sorted_values))[0:print_num])


# gen_response()