公司高端网站建设,wordpress无法点上传图片,手机网站营销的含义,昆明建设咨询监理有限公司网站Langchain 的 ConversationSummaryBufferMemory ConversationSummaryBufferMemory 在内存中保留最近交互的缓冲区#xff0c;但不仅仅是完全刷新旧的交互#xff0c;而是将它们编译成摘要并使用两者。但与之前的实现不同的是#xff0c;它使用令牌长度而不是交互次数来确定何… Langchain 的 ConversationSummaryBufferMemory ConversationSummaryBufferMemory 在内存中保留最近交互的缓冲区但不仅仅是完全刷新旧的交互而是将它们编译成摘要并使用两者。但与之前的实现不同的是它使用令牌长度而不是交互次数来确定何时刷新交互。
我们首先来了解一下如何使用这些实用程序
示例代码
from langchain.memory import ConversationSummaryBufferMemory
from langchain.llms import OpenAIllm OpenAI()memory ConversationSummaryBufferMemory(llmllm, max_token_limit40, return_messagesTrue
)
memory.save_context({input: 嗨}, {output: 最近怎么样})
memory.save_context({input: 没什么特别的你呢}, {output: 没什么特别的。})memory.load_memory_variables({})输出结果
{history: [SystemMessage(content\nThe human and AI are engaging in conversation. The human greets the AI and the AI asks how the human is doing., additional_kwargs{}),HumanMessage(content没什么特别的你呢, additional_kwargs{}, exampleFalse),AIMessage(content没什么特别的。, additional_kwargs{}, exampleFalse)]}示例代码
from langchain.chains import ConversationChainconversation_with_summary ConversationChain(llmllm,# We set a very low max_token_limit for the purposes of testing.memoryConversationSummaryBufferMemory(llmOpenAI(), max_token_limit250),verboseTrue,
)
conversation_with_summary.predict(input嗨最近怎么样)输出结果 Entering new ConversationChain chain...
Prompt after formatting:
The following is a friendly conversation between a human and an AI. The AI is talkative and provides lots of specific details from its context. If the AI does not know the answer to a question, it truthfully says it does not know.Current conversation:Human: 嗨最近怎么样
AI: Finished chain.嗨最近还不错我在学习新的技能并且正在尝试新的任务。我也在尝试改进我的语言处理能力以便更好地与人交流。你呢示例代码
conversation_with_summary.predict(input只是在写一些文档)输出结果
Entering new ConversationChain chain...
Prompt after formatting:
The following is a friendly conversation between a human and an AI. The AI is talkative and provides lots of specific details from its context. If the AI does not know the answer to a question, it truthfully says it does not know.Current conversation:
Human: 嗨最近怎么样
AI: 嗨最近还不错我在学习新的技能并且正在尝试新的任务。我也在尝试改进我的语言处理能力以便更好地与人交流。你呢
Human: 只是在写一些文档
AI: Finished chain.哦看起来很有趣你在写什么文档示例代码
conversation_with_summary.predict(input你听说过 LangChain 吗)输出结果 Entering new ConversationChain chain...
Prompt after formatting:
The following is a friendly conversation between a human and an AI. The AI is talkative and provides lots of specific details from its context. If the AI does not know the answer to a question, it truthfully says it does not know.Current conversation:
Human: 嗨最近怎么样
AI: 嗨最近还不错我在学习新的技能并且正在尝试新的任务。我也在尝试改进我的语言处理能力以便更好地与人交流。你呢
Human: 只是在写一些文档
AI: 哦看起来很有趣你在写什么文档
Human: 你听说过 LangChain 吗
AI: Finished chain.是的我知道LangChain。它是一个基于区块链的语言学习平台旨在帮助人们更好地学习外语。你正在写关于LangChain的文档吗示例代码
# 我们可以看到摘要和缓冲区都已更新
conversation_with_summary.predict(input哈哈不对不过很多人都把它混淆了。
)输出结果 Entering new ConversationChain chain...
Prompt after formatting:
The following is a friendly conversation between a human and an AI. The AI is talkative and provides lots of specific details from its context. If the AI does not know the answer to a question, it truthfully says it does not know.Current conversation:
System:
The AI thinks artificial intelligence is a force for good because it will help humans reach their full potential. It has been learning new skills and trying new tasks, as well as attempting to improve its language processing so that it can communicate better with humans. The human asks how the AI has been doing recently.
Human: 只是在写一些文档
AI: 哦看起来很有趣你在写什么文档
Human: 你听说过 LangChain 吗
AI: 是的我知道LangChain。它是一个基于区块链的语言学习平台旨在帮助人们更好地学习外语。你正在写关于LangChain的文档吗
Human: 哈哈不对不过很多人都把它混淆了。
AI: Finished chain.哦我明白了。那你在写什么文档呢完结
