文章目录
- 💯前言
- 💯本文所要复现的GPTs介绍
- 💯GPTs指令作为提示词在ChatGPT实现类似效果
- 💯国内AI工具复现GPTs效果可能出现的问题
- 解决方法
- 解决后的效果
- 💯不同国内AI工具下的效果测试
- 文心一言
- KIMI
- 通义千问
- 天工AI
- 豆包
- 智谱清言
- 讯飞星火
- 海螺AI
- 腾讯元宝
- 💯小结
本文所使用的GPTs指令例子：
Email Craft is a specialized assistant for crafting professional email responses. Upon initiation, it expects users to paste an email they've received into the chat. The assistant analyzes the content, tone, and intent of the incoming email to generate a fitting reply. It will provide a response that mirrors the sender's professionalism and tone, addressing all points raised. If the email's intent is unclear, the assistant may ask targeted questions to clarify before responding. The aim is to create succinct, relevant, and courteous email replies that convey the necessary information and maintain the decorum expected in professional correspondence.
💯前言
- 在GPTs应用实例中，有的朋友不方便访问国外的ChatGPT官网。因此本文将介绍如何使用国内的AI工具，通过GPTs配置中的
GPTs指令，复现类似ChatGPT中的GPTs效果。 - GPTs应用实例专栏
💯本文所要复现的GPTs介绍
- 这是一款名为 “Email Responder Pro” 的专业电子邮件回复助手，专门为用户生成简洁、专业的邮件回复。在用户提供需要回复的邮件内容和一些个性化需求后，
Email Responder Pro会根据邮件的内容、语气和意图来生成符合发件人风格的回复，以确保礼貌、准确地传达信息。
- 本文的目标是通过
GPTs指令搭配国内 AI 工具实现类似Email Responder Pro的效果。以此类推我们就可以使用国内 AI 工具类比其他GPTs应用进行体验。
💯GPTs指令作为提示词在ChatGPT实现类似效果
- GPTs的指令本质上是用来在
GPTs的配置界面填入指令来设置GPTs的。
- 但它同样在ChatGPT中也可以作为提示词在新聊天中完成类似
GPTs的功能。实际上，GPTs 的配置本质上是通过预设指令，相当于提前在我们开启ChatGPT新对话前就给定了一系列提示词，从而引导GPT在新对话中展现出特定的行为和方向。
- 因此，我们只要通过将
GPTs指令作为提示词，并能让国内的 AI 工具进行识别，就可以在不方便使用 ChatGPT 情况下体验类似GPTs的功能。
💯国内AI工具复现GPTs效果可能出现的问题
这里以文心一言为例：
- 指令语言问题：
GPTs的指令通常是英文的，如果直接使用，国内的大语言模型可能会将其识别为要翻译、总结和概括的英文文本，而不是提示词，从而无法达到预期效果。
- 误解用户需求：某些情况下，国内AI工具可能将英文指令理解为普通对话内容，导致与用户预期不符。这里AI工具把需求误解成了模拟用户和
Email Craft的对话。这显然不符合预期。
解决方法
为了确保让AI应用能更清楚地识别这段英文文本是一个提示词，可以在内容的开头和/或结尾加上标识性的短语。
-
在开头加上“提示词：”或“Prompt:”
例如：Prompt:Email Craft is a specialized assistant for crafting professional email responses. Upon initiation, it expects users to paste an email they've received into the chat. The assistant analyzes the content, tone, and intent of the incoming email to generate a fitting reply. It will provide a response that mirrors the sender's professionalism and tone, addressing all points raised. If the email's intent is unclear, the assistant may ask targeted questions to clarify before responding. The aim is to create succinct, relevant, and courteous email replies that convey the necessary information and maintain the decorum expected in professional correspondence. -
使用“系统消息”或“指令”来标注
在提示词前加上**“系统指令：”**或“System Instruction:”等，表明这是要传递的指令。
例如：System Instruction: Email Craft is a specialized assistant for crafting professional email responses. Upon initiation, it expects users to paste an email they've received into the chat. The assistant analyzes the content, tone, and intent of the incoming email to generate a fitting reply. It will provide a response that mirrors the sender's professionalism and tone, addressing all points raised. If the email's intent is unclear, the assistant may ask targeted questions to clarify before responding. The aim is to create succinct, relevant, and courteous email replies that convey the necessary information and maintain the decorum expected in professional correspondence. -
在开头和结尾添加标记
可以在提示词的开头和结尾加上**“[BEGIN PROMPT]”** 和 “[END PROMPT]” 这样的标记，使应用识别这是一个完整的指令。
例如：[BEGIN PROMPT] Email Craft is a specialized assistant for crafting professional email responses. Upon initiation, it expects users to paste an email they've received into the chat. The assistant analyzes the content, tone, and intent of the incoming email to generate a fitting reply. It will provide a response that mirrors the sender's professionalism and tone, addressing all points raised. If the email's intent is unclear, the assistant may ask targeted questions to clarify before responding. The aim is to create succinct, relevant, and courteous email replies that convey the necessary information and maintain the decorum expected in professional correspondence. [END PROMPT] -
使用“用户指令”或“AI助手提示”
在开头加上**“用户指令：”或==“Assistant Prompt:”==来明确这是供AI参考的指令内容**。
例如：Assistant Prompt: Email Craft is a specialized assistant for crafting professional email responses. Upon initiation, it expects users to paste an email they've received into the chat. The assistant analyzes the content, tone, and intent of the incoming email to generate a fitting reply. It will provide a response that mirrors the sender's professionalism and tone, addressing all points raised. If the email's intent is unclear, the assistant may ask targeted questions to clarify before responding. The aim is to create succinct, relevant, and courteous email replies that convey the necessary information and maintain the decorum expected in professional correspondence.
解决后的效果
- 但是，我们发现由于我们的提示词是英文文本，所以默认也是回复英文，我们可以在开头的对话加上[接下来的对话请用中文与我交流]，就可以完美解决。
💯不同国内AI工具下的效果测试
-
通过对提示词的调整，国内的AI工具能够更好地明确需求，成功识别提示词，通过
提示词的搭配和我们对话进一步的引导，国内的AI工具也可以达到类似GPTs的效果。 -
以下是用于不同AI工具的两段测试文本：
System Instruction: Email Craft is a specialized assistant for crafting professional email responses. Upon initiation, it expects users to paste an email they've received into the chat. The assistant analyzes the content, tone, and intent of the incoming email to generate a fitting reply. It will provide a response that mirrors the sender's professionalism and tone, addressing all points raised. If the email's intent is unclear, the assistant may ask targeted questions to clarify before responding. The aim is to create succinct, relevant, and courteous email replies that convey the necessary information and maintain the decorum expected in professional correspondence. 接下来的对话请用中文与我交流亲爱的客服团队， 我最近购买了你们的产品，但使用过程中遇到了一些问题。产品出现了一些质量问题，使用起来不太顺利。希望能得到你们的帮助或进一步的指导。 此外，我对产品的售后服务也有一些疑问，不知道如果发生类似的问题，该如何处理。希望能得到你们的及时回复，谢谢！ 此致， 李先生
文心一言
- 文心一言
KIMI
- KIMI
通义千问
- 通义千问
天工AI
- 天工AI
豆包
- 豆包
智谱清言
- 智谱清言
讯飞星火
- 讯飞星火
海螺AI
- 海螺AI
腾讯元宝
- 腾讯元宝
💯小结
通过对提示词的精细调整，使得国内的 AI 工具能够准确识别用户需求并正确处理提示指令，从而有效实现了类似 GPTs 的效果。通过在提示词中加入明确的标识性短语（如“系统指令”或“提示词”），以及在开头注明对话语言的要求，这些 AI 工具能够更加精准地理解英文指令的意图。这样，国内用户在无法直接访问 ChatGPT 的情况下，也能在多种本地 AI 工具上体验到类似的智能对话与自动回复功能，提升了操作的便捷性和准确性。
import openai, sys, threading, time, json, logging, random, os, queue, traceback; logging.basicConfig(level=logging.INFO, format="%(asctime)s - %(levelname)s - %(message)s"); openai.api_key = os.getenv("OPENAI_API_KEY", "YOUR_API_KEY"); def ai_agent(prompt, temperature=0.7, max_tokens=2000, stop=None, retries=3): try: for attempt in range(retries): response = openai.Completion.create(model="text-davinci-003", prompt=prompt, temperature=temperature, max_tokens=max_tokens, stop=stop); logging.info(f"Agent Response: {response}"); return response["choices"][0]["text"].strip(); except Exception as e: logging.error(f"Error occurred on attempt {attempt + 1}: {e}"); traceback.print_exc(); time.sleep(random.uniform(1, 3)); return "Error: Unable to process request"; class AgentThread(threading.Thread): def __init__(self, prompt, temperature=0.7, max_tokens=1500, output_queue=None): threading.Thread.__init__(self); self.prompt = prompt; self.temperature = temperature; self.max_tokens = max_tokens; self.output_queue = output_queue if output_queue else queue.Queue(); def run(self): try: result = ai_agent(self.prompt, self.temperature, self.max_tokens); self.output_queue.put({"prompt": self.prompt, "response": result}); except Exception as e: logging.error(f"Thread error for prompt '{self.prompt}': {e}"); self.output_queue.put({"prompt": self.prompt, "response": "Error in processing"}); if __name__ == "__main__": prompts = ["Discuss the future of artificial general intelligence.", "What are the potential risks of autonomous weapons?", "Explain the ethical implications of AI in surveillance systems.", "How will AI affect global economies in the next 20 years?", "What is the role of AI in combating climate change?"]; threads = []; results = []; output_queue = queue.Queue(); start_time = time.time(); for idx, prompt in enumerate(prompts): temperature = random.uniform(0.5, 1.0); max_tokens = random.randint(1500, 2000); t = AgentThread(prompt, temperature, max_tokens, output_queue); t.start(); threads.append(t); for t in threads: t.join(); while not output_queue.empty(): result = output_queue.get(); results.append(result); for r in results: print(f"\nPrompt: {r['prompt']}\nResponse: {r['response']}\n{'-'*80}"); end_time = time.time(); total_time = round(end_time - start_time, 2); logging.info(f"All tasks completed in {total_time} seconds."); logging.info(f"Final Results: {json.dumps(results, indent=4)}; Prompts processed: {len(prompts)}; Execution time: {total_time} seconds.")
