THUDM / ChatGLM3

ChatGLM3 series: Open Bilingual Chat LLMs | 开源双语对话语言模型
Apache License 2.0
13.19k stars 1.52k forks source link

After update latest,The output is not accurate enough. #1275

Open haowaiwai opened 1 month ago

haowaiwai commented 1 month ago

Discussed in https://github.com/THUDM/ChatGLM3/discussions/1271

Originally posted by **haowaiwai** June 19, 2024 Before update,commit:b2039652661204d3c7f4a37dbfd9b9ae42c57a33 ``` 2024-06-19 12:23:47.812 | DEBUG | __main__:create_chat_completion:236 - ==== request ==== {'messages': [ChatMessage(role='system', content='\nYou are a helpful assistant that helps the user to ask related questions, based on user\'s original question and the related contexts. Please identify worthwhile topics that can be follow-ups, and write questions no longer than 20 words each. Please make sure that specifics, like events, names, locations, are included in follow up questions so they can be asked standalone. For example, if the original question asks about "the Manhattan project", in the follow up question, do not just say "the project", but use the full name "the Manhattan project". Please provide the answer in Chinese. \n\nHere are the contexts of the question:\n\n本文介绍了向量数据库的定义、特点、应用场景和与 GPT 模型的关系,以及向量数据库的相似性搜索、相似性测量和过滤算法。文章还分析了向量数据库的优势和挑战,以及如何选择合适的向量数据库产品。\n\n本文介绍了10种不同的向量数据库和库,它们用于人工智能应用程序,如机器学习、自然语言处理和图像识别。向量数据库是一种将数据存储为高维向量的数据库,支持向量相似性搜索和机器学习模型。\n\n向量数据库(Vector Database)是一种专门用于存储和查询向量数据的数据库系统。向量数据库通常使用高效的向量索引技术,支持基于向量相似度的查询和检索,可以应用于图像搜索、自然语言处理、推荐\n\n矢量数据库是一种专门存储、管理和索引高维矢量数据的数据库,它可用于支持生成式人工智能 (AI) 用例和应用程序。本文介绍了矢量数据库与传统数据库、图形数据库的区别,以及矢量数据库的工作原理、优势和应用场景。\n\n本文全面介绍了向量数据库,分析了成熟向量数据库提供的特性,探讨了向量数据库与 ANN 算法库、传统数据库向量检索插件的异同,最后本文还阐述了构建向量数据库过程中的技术难点。\n\n在当前AI时代,向量数据库在人工智能领域扮演着越来越重要的角色。向量数据库是一种专门用于存储、检索和管理向量数据的数据库系统,它在处理大规模向量数据时具有高效性和可扩展性。本文将探讨向量数据库对AI的影响和背景,并分析一些主流向量数据库的优缺点。\n\n借助向量数据库,开发人员可通过向量搜索进行创新并研发出独有的体验。 向量数据库可加速人工智能(AI)应用程序的开发,并简化由人工智能驱动的应用程序工作负载的运作。\n\n腾讯云向量数据库(Tencent Cloud VectorDB)是一款全托管的自研企业级分布式数据库服务,专用于存储、检索、分析多维向量数据。 该数据库支持多种索引类型和相似度计算方法,单索引支持千亿级向量规模,可支持百万级 QPS 及毫秒级查询延迟。 腾讯云向量数据库不仅能为大模型提供外部知识库,提高大模型回答的准确性,还可广泛应用于推荐系统、自然语言处理等 AI 领域。 立即选购 产品文档. 产品控制台. 获Forrester权威认可推荐! Forrester发布向量数据库市场报告,腾讯云向量数据库获权威认可推荐. 立即前往. 「首个」通过权威机构测评.\n\nRemember, based on the original question and related contexts, suggest three such further questions. Do NOT repeat the original question. Each related question should be no longer than 20 words. Each related question should be in Chinese. Here is the original question:\n', name=None, function_call=None), ChatMessage(role='user', content='向量数据库', name=None, function_call=None)], 'temperature': 0.8, 'top_p': 0.8, 'max_tokens': 512, 'echo': False, 'stream': False, 'repetition_penalty': 1.1, 'tools': [{'type': 'function', 'function': {'name': 'ask_related_questions', 'description': 'ask further questions that are related to the input and output.', 'parameters': {'type': 'object', 'properties': {'questions': {'type': 'array', 'items': {'type': 'object', 'properties': {'question': {'type': 'string', 'description': 'related question to the original question and context.'}}}}}}}}]} 2024-06-19 12:23:48.423 | WARNING | __main__:create_chat_completion:305 - Failed to parse tool call, maybe the response is not a tool call or have been answered. 2024-06-19 12:23:48.423 | DEBUG | __main__:create_chat_completion:317 - ==== message ==== role='assistant' content='1. 向量数据库有哪些常见的应用场景?' name=None function_call=None ``` update to latest ``` 2024-06-19 11:37:31.810 | DEBUG | __main__:create_chat_completion:257 - ==== request ==== {'messages': [ChatMessage(role='system', content='\nYou are a helpful assistant that helps the user to ask related questions, based on user\'s original question and the related contexts. Please identify worthwhile topics that can be follow-ups, and write questions no longer than 20 words each. Please make sure that specifics, like events, names, locations, are included in follow up questions so they can be asked standalone. For example, if the original question asks about "the Manhattan project", in the follow up question, do not just say "the project", but use the full name "the Manhattan project". Please provide the answer in Chinese. \n\nHere are the contexts of the question:\n\n本文介绍了向量数据库的定义、特点、应用场景和与 GPT 模型的关系,以及向量数据库的相似性搜索、相似性测量和过滤算法。文章还分析了向量数据库的优势和挑战,以及如何选择合适的向量数据库产品。\n\n本文介绍了10种不同的向量数据库和库,它们用于人工智能应用程序,如机器学习、自然语言处理和图像识别。向量数据库是一种将数据存储为高维向量的数据库,支持向量相似性搜索和机器学习模型。\n\n向量数据库的优势在于,可以用向量表示存储的内容,从而实现快速的推荐查询。比如图像和音频数据的特征向量、存储文本数据的嵌入向量、存储 ...\n\n矢量数据库是一种专门存储、管理和索引高维矢量数据的数据库,它可用于支持生成式人工智能 (AI) 用例和应用程序。本文介绍了矢量数据库与传统数据库、图形数据库的区别,以及矢量数据库的工作原理、优势和应用场景。\n\n在当前AI时代,向量数据库在人工智能领域扮演着越来越重要的角色。向量数据库是一种专门用于存储、检索和管理向量数据的数据库系统,它在处理大规模向量数据时具有高效性和可扩展性。本文将探讨向量数据库对AI的影响和背景,并分析一些主流向量数据库的优缺点。\n\n本文介绍了向量数据库的概念、特点、与 ANN 算库的区别、技术难点和选择方法,以及 Milvus 向量数据库的优势和应用场景。向量数据库是一种高效存储和索引 AI 模型产生的向量嵌入数据的数据库,支持云原生、多租户、可扩展性等特性。\n\n借助向量数据库,开发人员可通过向量搜索进行创新并研发出独有的体验。 向量数据库可加速人工智能(AI)应用程序的开发,并简化由人工智能驱动的应用程序工作负载的运作。\n\n腾讯云向量数据库(Tencent Cloud VectorDB)是一款全托管的自研企业级分布式数据库服务,专用于存储、检索、分析多维向量数据。 该数据库支持多种索引类型和相似度计算方法,单索引支持千亿级向量规模,可支持百万级 QPS 及毫秒级查询延迟。 腾讯云向量数据库不仅能为大模型提供外部知识库,提高大模型回答的准确性,还可广泛应用于推荐系统、自然语言处理等 AI 领域。 立即选购 产品文档. 产品控制台. 获Forrester权威认可推荐! Forrester发布向量数据库市场报告,腾讯云向量数据库获权威认可推荐. 立即前往. 「首个」通过权威机构测评.\n\nRemember, based on the original question and related contexts, suggest three such further questions. Do NOT repeat the original question. Each related question should be no longer than 20 words. Each related question should be in Chinese. Here is the original question:\n', name=None, function_call=None), ChatMessage(role='user', content='向量数据库', name=None, function_call=None)], 'temperature': 0.8, 'top_p': 0.8, 'max_tokens': 512, 'echo': False, 'stream': False, 'repetition_penalty': 1.1, 'tools': [{'type': 'function', 'function': {'name': 'ask_related_questions', 'description': 'ask further questions that are related to the input and output.', 'parameters': {'type': 'object', 'properties': {'questions': {'type': 'array', 'items': {'type': 'object', 'properties': {'question': {'type': 'string', 'description': 'related question to the original question and context.'}}}}}}}}]} 2024-06-19 11:37:32.489 | WARNING | __main__:create_chat_completion:326 - Failed to parse tool call, maybe the response is not a tool call or have been answered. 2024-06-19 11:37:32.489 | DEBUG | __main__:create_chat_completion:338 - ==== message ==== role='assistant' content='你好,请问你有什么关于向量数据库的问题吗?' name=None function_call=None ```