Semantic Kernel 入门系列：🥑Memory内存
了解的运作原理之后，就可以开始使用Semantic Kernel来制作应用了。
Semantic Kernel将embedding的功能封装到了Memory中，用来存储上下文信息，就好像电脑的内存一样，而LLM就像是CPU一样，我们所需要做的就是从内存中取出相关的信息交给CPU处理就好了。
Memory配置
首先需要安装Microsoft.SemanticKernel.Plugins.Memory NuGet包,Memory的功能作为Plugin提供。
使用Memory需要注册 embedding模型，目前常用的就是 text-embedding-ada-002。同时需要添加MemoryStore，用于存储更多的信息，这里Semantic Kernel提供了一个 VolatileMemoryStore，就是一个普通的内存存储的MemoryStore，可以用作测试，实际生产使用的话，建议采用其他的向量数据库的存储。
var memory = new MemoryBuilder()
.WithAzureOpenAITextEmbeddingGeneration("text-embedding-ada-002",
Environment.GetEnvironmentVariable("AZURE_OPENAI_API_ENDPOINT"),
Environment.GetEnvironmentVariable("AZURE_OPENAI_API_KEY"))
.WithMemoryStore(new VolatileMemoryStore())
.Build();
信息存储
完成了基础信息的注册后，就可以往Memory中存储信息了。
const string MemoryCollectionName = "aboutMe";
await memory.SaveInformationAsync(MemoryCollectionName, id: "info1", text: "My name is Andrea");
await memory.SaveInformationAsync(MemoryCollectionName, id: "info2", text: "I currently work as a tourist operator");
await memory.SaveInformationAsync(MemoryCollectionName, id: "info3", text: "I currently live in Seattle and have been living there since 2005");
await memory.SaveInformationAsync(MemoryCollectionName, id: "info4", text: "I visited France and Italy five times since 2015");
await memory.SaveInformationAsync(MemoryCollectionName, id: "info5", text: "My family is from New York");
SaveInformationAsync 会将text的内容通过 embedding 模型转化为对应的文本向量，存放在的MemoryStore中。其中CollectionName如同数据库的表名，Id就是Id。
语义搜索
完成信息的存储之后，就可以用来语义搜索了。
直接使用Memory.SearchAsync方法，指定对应的Collection，同时提供相应的查询问题，查询问题也会被转化为embedding，再在MemoryStore中计算查找最相似的信息。
var questions = new[]
{
"what is my name?",
"where do I live?",
"where is my family from?",
"where have I travelled?",
"what do I do for work?",
};
foreach (var q in questions)
{
await foreach (var response in memory.SearchAsync(MemoryCollectionName, q))
{
Console.WriteLine(q + " " + response?.Metadata.Text);
}
}
// output
/*
what is my name? My name is Andrea
where do I live? I currently live in Seattle and have been living there since 2005
where is my family from? My family is from New York
where have I travelled? I visited France and Italy five times since 2015
what do I do for work? I currently work as a tourist operator
*/
到这个时候，即便不需要进行总结归纳，光是这样的语义查找，都会很有价值。
引用存储
除了添加信息以外，还可以添加引用，像是非常有用的参考链接之类的。
const string memoryCollectionName = "SKGitHub";
var githubFiles = new Dictionary<string, string>()
{
["https://github.com/microsoft/semantic-kernel/blob/main/README.md"]
= "README: Installation, getting started, and how to contribute",
["https://github.com/microsoft/semantic-kernel/blob/main/samples/notebooks/dotnet/2-running-prompts-from-file.ipynb"]
= "Jupyter notebook describing how to pass prompts from a file to a semantic plugin or function",
["https://github.com/microsoft/semantic-kernel/blob/main/samples/notebooks/dotnet/Getting-Started-Notebook.ipynb"]
= "Jupyter notebook describing how to get started with the Semantic Kernel",
["https://github.com/microsoft/semantic-kernel/tree/main/samples/plugins/Chatplugin/ChatGPT"]
= "Sample demonstrating how to create a chat plugin interfacing with ChatGPT",
["https://github.com/microsoft/semantic-kernel/blob/main/dotnet/src/SemanticKernel/Memory/Volatile/VolatileMemoryStore.cs"]
= "C# class that defines a volatile embedding store",
["https://github.com/microsoft/semantic-kernel/tree/main/samples/dotnet/KernelHttpServer/README.md"]
= "README: How to set up a Semantic Kernel Service API using Azure Function Runtime v4",
["https://github.com/microsoft/semantic-kernel/tree/main/samples/apps/chat-summary-webapp-react/README.md"]
= "README: README associated with a sample starter react-based chat summary webapp",
};
Console.WriteLine("Adding some GitHub file URLs and their descriptions to a volatile Semantic Memory.");
foreach (var entry in githubFiles)
{
await memory.SaveReferenceAsync(
collection: memoryCollectionName,
description: entry.Value,
text: entry.Value,
externalId: entry.Key,
externalSourceName: "GitHub"
);
}
同样的，使用SearchAsync搜索就行。
string ask = "I love Jupyter notebooks, how should I get started?";
Console.WriteLine("===========================\n" +
"Query: " + ask + "\n");
var memories = memory.SearchAsync(memoryCollectionName, ask, limit: 5, minRelevanceScore: 0.77);
var i = 0;
await foreach (MemoryQueryResult result in memories)
{
Console.WriteLine($"Result {++i}:");
Console.WriteLine(" URL: : " + result.Metadata.Id);
Console.WriteLine(" Title : " + result.Metadata.Description);
Console.WriteLine(" ExternalSource: " + result.Metadata.ExternalSourceName);
Console.WriteLine(" Relevance: " + result.Relevance);
Console.WriteLine();
}
//output
/*
===========================
Query: I love Jupyter notebooks, how should I get started?
Result 1:
URL: : https://github.com/microsoft/semantic-kernel/blob/main/samples/notebooks/dotnet/Getting-Started-Notebook.ipynb
Title : Jupyter notebook describing how to get started with the Semantic Kernel
ExternalSource: GitHub
Relevance: 0.8677417635917664
Result 2:
URL: : https://github.com/microsoft/semantic-kernel/blob/main/samples/notebooks/dotnet/2-running-prompts-from-file.ipynb
Title : Jupyter notebook describing how to pass prompts from a file to a semantic plugin or function
ExternalSource: GitHub
Relevance: 0.8165638446807861
Result 3:
URL: : https://github.com/microsoft/semantic-kernel/blob/main/README.md
Title : README: Installation, getting started, and how to contribute
ExternalSource: GitHub
Relevance: 0.808631956577301
*/
这里多使用了两个参数，一个是limit，用于限制返回信息的条数,只返回最相似的前几条数据，另外一个是minRelevanceScore，限制最小的相关度分数，这个取值范围在0.0 ~ 1.0 之间，1.0意味着完全匹配。
语义问答
将Memory的存储、搜索功能和语义插件相结合，就可以快速的打造一个实用的语义问答的应用了。
只需要将搜索到的相关信息内容填充到 prompt中，然后将内容和问题都抛给LLM，就可以等着得到一个满意的答案了。
var prompt =
"""
It can give explicit instructions or say 'I don't know' if it does not have an answer.
Information about me, from previous conversations:
{{ $fact }}
User: {{ $ask }}
ChatBot:
""";
var function = kernel.CreateFunctionFromPrompt(prompt);
var ask = "Hello, I think we've met before, remember? my name is...";
var fact = await memory.SearchAsync(MemoryCollectionName, ask).FirstOrDefaultAsync(); // .FirstOrDefaultAsync() 需要安装 System.Linq.Async
var args = new KernelArguments();
args["fact"] = fact?.Metadata?.Text;
args["ask"] = ask;
var resultContext = await function.InvokeAsync(kernel,args);
resultContext.GetValue<string>().Dump();
//output
/*
Hi there! Yes, I remember you. Your name is Andrea, right?
*/
优化搜索过程
由于这种场景太常见了，所以Semantic Kernel中直接提供了一个插件TextMemoryPlugin，通过Function调用的方式简化了搜索的过程。
// .. SaveInformations
// TextMemoryPlugin provides the "recall" function
kernel.ImportPluginFromObject(new TextMemoryPlugin(memory));
var prompt =
"""
It can give explicit instructions or say 'I don't know' if it does not have an answer.
Information about me, from previous conversations:
{{ recall $ask }}
User: {{ $ask }}
ChatBot:
""";
var function = kernel.CreateFunctionFromPrompt(prompt);
var ask = "Hello, I think we've met before, remember? my name is...";
var args = new KernelArguments();
args["ask"] = ask;
args[TextMemoryPlugin.CollectionParam] = MemoryCollectionName;
var resultContext = await function.InvokeAsync(kernel, args);
resultContext.GetValue<string>().Dump();
// output
/*
Hi there! Yes, I remember you. Your name is Andrea, right?
*/
这里直接使用 recall 方法，将问题传给了 TextMemoryPlugin，搜索对应得到结果，免去了手动搜索注入得过程。
内存的持久化
VolatileMemoryStore本身也是易丢失的，往往使用到内存的场景，其中的信息都是有可能长期存储的，起码并不会即刻过期。那么将这些信息的 embedding 能够长期存储起来，也是比较划算的事情。毕竟每一次做 embedding的转化也是需要调接口，需要花钱的。
Semantic Kernel库中包含了SQLite、Qdrant和CosmosDB的实现，自行扩展的话，也只需要实现 IMemoryStore 这个接口就可以了。
参考资料：
