1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16

curl -fsSL https://ollama.com/install.sh | sh

>>> Installing ollama to /usr/local
>>> Downloading Linux amd64 CLI
######################################################################## 100.0%#=#=-#  #                                                                  ######################################################################## 100.0%
>>> Making ollama accessible in the PATH in /usr/local/bin
>>> Creating ollama user...
>>> Adding ollama user to render group...
>>> Adding ollama user to video group...
>>> Adding current user to ollama group...
>>> Creating ollama systemd service...
>>> Enabling and starting ollama service...
Created symlink /etc/systemd/system/default.target.wants/ollama.service → /etc/systemd/system/ollama.service.
>>> The Ollama API is now available at 127.0.0.1:11434.
>>> Install complete. Run "ollama" from the command line.
WARNING: No NVIDIA/AMD GPU detected. Ollama will run in CPU-only mode.

1
2

✗ curl 127.0.0.1:11434 
Ollama is running%                   

1

ollama pull llama3

1

ollama rm llama3

1

ollama cp llama3 my-model

1
2

>>> What's in this image? /Users/jmorgan/Desktop/smile.png
The image features a yellow smiley face, which is likely the central focus of the picture.

1
2

$ ollama run llama3 "Summarize this file: $(cat README.md)"
 Ollama is a lightweight, extensible framework for building and running language models on the local machine. It provides a simple API for creating, running, and managing models, as well as a library of pre-built models that can be easily used in a variety of applications.

1
2
3
4

curl http://localhost:11434/api/generate -d '{
  "model": "llama3",
  "prompt":"Why is the sky blue?"
}'

1
2
3
4
5
6

curl http://localhost:11434/api/chat -d '{
  "model": "llama3",
  "messages": [
    { "role": "user", "content": "why is the sky blue?" }
  ]
}'

1
2
3
4

git clone https://github.com/ollama-webui/ollama-webui-lite.git
cd ollama-webui-lite
yarn
yarn dev

1
2

docker pull qdrant/qdrant
docker run -itd --name qdrant -p 6333:6333 qdrant/qdrant

1
2
3
4
5
6
7
8
9

curl -X PUT http://localhost:6333/collections/langchaingo-ollama-rag \
  -H 'Content-Type: application/json' \
  --data-raw '{
    "vectors": {
      "size": 768,
      "distance": "Dot"
    }
  }'

1

curl --location --request DELETE 'http://localhost:6333/collections/langchaingo-ollama-rag'

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21

// TextToChunks 函数将文本文件转换为文档块
func TextToChunks(dirFile string, chunkSize, chunkOverlap int) ([]schema.Document, error) {
	file, err := os.Open(dirFile)
	if err != nil {
		return nil, err
	}
	// 创建一个新的文本文档加载器
	docLoaded := documentloaders.NewText(file)
	// 创建一个新的递归字符文本分割器
	split := textsplitter.NewRecursiveCharacter()
	// 设置块大小
	split.ChunkSize = chunkSize
	// 设置块重叠大小
	split.ChunkOverlap = chunkOverlap
	// 加载并分割文档
	docs, err := docLoaded.LoadAndSplit(context.Background(), split)
	if err != nil {
		return nil, err
	}
	return docs, nil
}

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18

➜  langchaingo-ollama-rag go run *.go filetochunks -f livenet.txt 
INFO   [2024-08-25 22:06:04]main.init.func1:80 转换文件为块成功，块数量: 458                          
🗂 块儿内容==> LiveNet: A Low-Latency Video Transport Network for
Large-Scale Live Streaming
Jinyang Li† § ‡ , Zhenyu Li† ‡, Ri Lu§, Kai Xiao§, Songlin Li§, Jufeng Chen§, Jingyu Yang§,
🗂 块儿内容==> Chunli Zong§, Aiyun Chen§, Qinghua Wu† ‡, Chen Sun§, Gareth Tyson∗, Hongqiang Harry Liu§
†Institute of Computing Technology, Chinese Academy of Sciences §Alibaba Group
🗂 块儿内容==> ‡University of Chinese Academy of Sciences ∗Hong Kong University of Science and Technology (GZ)
ABSTRACT
Low-latency live streaming has imposed stringent latency require-
🗂 块儿内容==> ments on video transport networks. In this paper, we report on the
design and operation of the Alibaba low-latency video transport
network, LiveNet. LiveNet builds on a flat CDN overlay with a
🗂 块儿内容==> centralized controller for global optimization. As part of this, we
present our design of the global routing computation and path as-
signment, as well as our fast data transmission architecture with
🗂 块儿内容==> fine-grained control of video frames. The performance results ob-

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34

// storeDocs 将文档存储到向量数据库
func storeDocs(docs []schema.Document, store *qdrant.Store) error {
	// 如果文档数组长度大于0
	if len(docs) > 0 {
		// 添加文档到存储
		_, err := store.AddDocuments(context.Background(), docs)
		if err != nil {
			return err
		}
	}
	return nil
}

// useRetriaver 函数使用检索器
func useRetriaver(store *qdrant.Store, prompt string, topk int) ([]schema.Document, error) {
	// 设置选项向量
	optionsVector := []vectorstores.Option{
		vectorstores.WithScoreThreshold(0.80), // 设置分数阈值
	}

	// 创建检索器
	retriever := vectorstores.ToRetriever(store, topk, optionsVector...)
	// 搜索
	docRetrieved, err := retriever.GetRelevantDocuments(context.Background(), prompt)

	if err != nil {
		return nil, fmt.Errorf("检索文档失败: %v", err)
	}

	// 返回检索到的文档
	return docRetrieved, nil
}

1
2
3
4
5
6
7
8

$ go run *.go embedding                                                 
INFO   [2024-08-25 22:57:28]main.init.func3:101 转换块为向量成功 
$ go run *.go retriever -t 3
请输入你的问题: CDN
🗂 根据输入的内容检索出的块内容==> mapped to CDN servers via DNS redirection. The CDN node that a broadcaster connects to is called a producer node, and the CDN node that a viewer connects to is called a consumer node. A broad-
🗂 根据输入的内容检索出的块内容==> Dec. 10 Dec. 11-12 Dec. 13 CDN path delay (ms) 188 192 180 CDN path length 2 2 2 Streaming delay (ms) 954 988 944 0-stall ratio (%) 97 97 97 Fast startup ratio (%) 94 94 95
🗂 根据输入的内容检索出的块内容==> Network control algorithms; KEYWORDS CDN; Low latency transmission; Live streaming ACM Reference Format: Jinyang Li† § ‡ , Zhenyu Li† ‡, Ri Lu§, Kai Xiao§, Songlin Li§, Jufeng Chen§,

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29

// GetAnswer 获取答案
func GetAnswer(ctx context.Context, llm llms.Model, docRetrieved []schema.Document, prompt string) (string, error) {
	// 创建一个新的聊天消息历史记录
	history := memory.NewChatMessageHistory()
	// 将检索到的文档添加到历史记录中
	for _, doc := range docRetrieved {
		history.AddAIMessage(ctx, doc.PageContent)
	}
	// 使用历史记录创建一个新的对话缓冲区
	conversation := memory.NewConversationBuffer(memory.WithChatHistory(history))

	executor := agents.NewExecutor(
		agents.NewConversationalAgent(llm, nil),
		nil,
		agents.WithMemory(conversation),
	)
	// 设置链调用选项
	options := []chains.ChainCallOption{
		chains.WithTemperature(0.8),
	}
	// 运行链
	res, err := chains.Run(ctx, executor, prompt, options...)
	if err != nil {
		return "", err
	}

	return res, nil
}

1

ollama run llama2-chinese

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12

// Translate 将文本翻译为中文
func Translate(llm llms.Model, text string) (string, error) {
	completion, err := llms.GenerateFromSinglePrompt(
		context.TODO(),
		llm,
		"将如下这句话翻译为中文，只需要回复翻译后的内容，而不需要回复其他任何内容。需要翻译的英文内容是: \n"+text,
		llms.WithTemperature(0.8))
	if err != nil {
		return "", err
	}
	return completion, nil
}

1
2
3
4
5

go run *.go getanswer
请输入你的问题: 导入的文章主要讲了什么?
🗂 原始回答==>  The article appears to be discussing the main topic of what the imported article is about.

🗂 翻译后的回答==> 这句话翻译为中文：关于引进报道主题的文章。