<?xml version="1.0" encoding="utf-8" ?><rss version="2.0"><channel><title><![CDATA[python_pygame_bl的博客]]></title><description><![CDATA[]]></description><link>https://blog.youkuaiyun.com/python_pygame_bl</link><language>zh-cn</language><generator>https://blog.youkuaiyun.com/</generator><copyright><![CDATA[Copyright &copy; python_pygame_bl]]></copyright><item><title><![CDATA[本地大模型实现长期记忆--让AI记住你的喜好（基于Langchain）]]></title><link>https://blog.youkuaiyun.com/python_pygame_bl/article/details/145806682</link><guid>https://blog.youkuaiyun.com/python_pygame_bl/article/details/145806682</guid><author>python_pygame_bl</author><pubDate>Sun, 23 Feb 2025 11:08:05 +0800</pubDate><description><![CDATA[通过langchain的方法可以实现模型的长期记忆，将对话历史向量化后保存在本地，以下是脚本。脚本会自动生成向量化后的记忆文件。]]></description><category></category></item><item><title><![CDATA[本地部署DeepSeek与Llama3，基于Langchain和Ollama实现本地双模型的交互对话]]></title><link>https://blog.youkuaiyun.com/python_pygame_bl/article/details/145780144</link><guid>https://blog.youkuaiyun.com/python_pygame_bl/article/details/145780144</guid><author>python_pygame_bl</author><pubDate>Fri, 21 Feb 2025 16:35:35 +0800</pubDate><description><![CDATA[1.创建三个python脚本，‘llm1.py’载入deepseek模型，‘llm2.py’载入llama3模型，‘main.py’用来同步运行前两个脚本。访问ollama官网https://ollama.com/，下载OllamaSetup.exe并安装，ollama会默认添加到环境变量中。2.在目录中创建‘llm1输出.txt’和‘llm2输出.txt’用来实现两个模型之间的对话读取，创建'log.txt'记录对话。为了获得更快的速度，可以安装cuda，让模型在GPU上运行。]]></description><category></category></item></channel></rss>