Python ChatGPT 机器人在 openai.Completion 中的问题

Question

我正在尝试写一个聊天机器人。结果出现了一个错误，提示“你试图访问 openai.Completion，但在 openai>=1.0.0 中这已经不再支持了。”

有人能帮我想办法把这个代码改写成可以在 openAI API 的 v1.0.0 版本中使用的吗？

import openai

# Replace 'your_api_key_here' with your actual OpenAI API key
openai.api_key = 'your_api_key_here'

def chat_with_gpt(prompt):
    """
    Sends a prompt to GPT and returns the generated response.
    
    :param prompt: The input text to send.
    :return: The text of the generated response.
    """
    try:
        response = openai.Completion.create(
            model="gpt-3.5-turbo",  # Adjust with the appropriate model you have access to
            prompt=prompt,
            temperature=0.7,
            max_tokens=150,
            n=1,
            stop=None
        )
        return response.choices[0].text.strip()
    except Exception as e:
        return f"An error occurred: {str(e)}"

def main():
    print("GPT-3.5: Hello! How can I assist you today? Type 'quit' to exit.")
    chat_history = []  # To store the conversation history

    while True:
        user_input = input("You: ")
        if user_input.lower() == "quit":
            print("GPT-3.5: Goodbye! Have a great day!")
            break
        
        # Concatenate the chat history with the new message for context
        prompt = "\n".join(chat_history + [f"You: {user_input}", "GPT-3.5:"])
        response = chat_with_gpt(prompt)
        
        print(f"GPT-3.5: {response}")
        
        # Update the chat history
        chat_history.extend([f"You: {user_input}", f"GPT-3.5: {response}"])
        # Optional: Limit the history size to the last few exchanges to manage token limits
        chat_history = chat_history[-6:]

if __name__ == "__main__":
    main()

我也试过 openai.ChatCompletion，但也遇到了类似的错误。

Answer 1

你不需要使用 openai.Completion.create() 这个方法，而是要先创建一个客户端对象，方法是用 client = openai.OpenAI()。然后，你可以使用 client.chat.completions.create() 这个方法来生成内容。

下面是一个示例，展示了你可以怎么做（这个例子来自官方的 OpenAI 文档）：

from openai import OpenAI
client = OpenAI()

completion = client.chat.completions.create(
  model="gpt-3.5-turbo",
  messages=[
    {"role": "system", "content": "You are a poetic assistant, skilled in explaining complex programming concepts with creative flair."},
    {"role": "user", "content": "Compose a poem that explains the concept of recursion in programming."}
  ]
)

print(completion.choices[0].message)