Class: Langchain::Vectorsearch::Milvus

Inherits:
Base
  • Object
show all
Defined in:
lib/langchain/vectorsearch/milvus.rb

Constant Summary

Constants inherited from Base

Base::DEFAULT_METRIC

Instance Attribute Summary

Attributes inherited from Base

#client, #index_name, #llm

Instance Method Summary collapse

Methods inherited from Base

#add_data, #generate_hyde_prompt, #generate_rag_prompt, #similarity_search_with_hyde, #update_texts

Methods included from DependencyHelper

#depends_on

Constructor Details

#initialize(url:, index_name:, llm:, api_key: nil) ⇒ Milvus

Wrapper around Milvus REST APIs.

Gem requirements:

gem "milvus", "~> 0.10.3"

Usage:

milvus = Langchain::Vectorsearch::Milvus.new(url:, index_name:, llm:, api_key:)



14
15
16
17
18
19
20
21
22
23
24
25
 
# File 'lib/langchain/vectorsearch/milvus.rb', line 14

def initialize(url:, index_name:, llm:, api_key: nil)
  depends_on "milvus"

  @client = ::Milvus::Client.new(
    url: url,
    api_key: api_key,
    logger: Langchain.logger
  )
  @index_name = index_name

  super(llm: llm)
end

Instance Method Details

#add_texts(texts:) ⇒ Object 



27
28
29
30
31
32
33
34
 
# File 'lib/langchain/vectorsearch/milvus.rb', line 27

def add_texts(texts:)
  client.entities.insert(
    collection_name: index_name,
    data: texts.map do |text|
      {content: text, vector: llm.embed(text: text).embedding}
    end
  )
end

#ask(question:, k: 4) {|String| ... } ⇒ String

Ask a question and return the answer

Parameters:

  • question (String) —

    The question to ask

  • k (Integer) (defaults to: 4) —

    The number of results to have in context

Yields:

  • (String) —

    Stream responses back one String at a time

Returns:

  • (String) —

    The answer to the question



145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
 
# File 'lib/langchain/vectorsearch/milvus.rb', line 145

def ask(question:, k: 4, &block)
  search_results = similarity_search(query: question, k: k)

  content_data = search_results.dig("data").map { |result| result.dig("content") }

  context = content_data.join("\n---\n")

  prompt = generate_rag_prompt(question: question, context: context)

  messages = [{role: "user", content: prompt}]
  response = llm.chat(messages: messages, &block)

  response.context = context
  response
end

#create_default_index ⇒ Boolean

Create the default index

Returns:

  • (Boolean) —

    The response from the server



85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
 
# File 'lib/langchain/vectorsearch/milvus.rb', line 85

def create_default_index
  client.indexes.create(
    collection_name: index_name,
    index_params: [
      {
        metricType: "L2",
        fieldName: "vector",
        indexName: "vector_idx",
        indexConfig: {
          index_type: "AUTOINDEX"
        }
      }
    ]
  )
end

#create_default_schema ⇒ Hash

Create default schema

Returns:

  • (Hash) —

    The response from the server



55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
 
# File 'lib/langchain/vectorsearch/milvus.rb', line 55

def create_default_schema
  client.collections.create(
    auto_id: true,
    collection_name: index_name,
    fields: [
      {
        fieldName: "id",
        isPrimary: true,
        dataType: "Int64"
      }, {
        fieldName: "content",
        isPrimary: false,
        dataType: "VarChar",
        elementTypeParams: {
          max_length: "32768" # Largest allowed value
        }
      }, {
        fieldName: "vector",
        isPrimary: false,
        dataType: "FloatVector",
        elementTypeParams: {
          dim: llm.default_dimensions.to_s
        }
      }
    ]
  )
end

#destroy_default_schema ⇒ Hash

Delete default schema

Returns:

  • (Hash) —

    The response from the server



109
110
111
 
# File 'lib/langchain/vectorsearch/milvus.rb', line 109

def destroy_default_schema
  client.collections.drop(collection_name: index_name)
end

#get_default_schema ⇒ Hash

Get the default schema

Returns:

  • (Hash) —

    The response from the server



103
104
105
 
# File 'lib/langchain/vectorsearch/milvus.rb', line 103

def get_default_schema
  client.collections.describe(collection_name: index_name)
end

#load_default_schema ⇒ Boolean

Load default schema into memory

Returns:

  • (Boolean) —

    The response from the server



115
116
117
 
# File 'lib/langchain/vectorsearch/milvus.rb', line 115

def load_default_schema
  client.collections.load(collection_name: index_name)
end

#remove_texts(ids:) ⇒ Boolean

Deletes a list of texts in the index

Parameters:

  • ids (Array<Integer>) —

    The ids of texts to delete

Returns:

  • (Boolean) —

    The response from the server

Raises:

  • (ArgumentError) 


42
43
44
45
46
47
48
49
 
# File 'lib/langchain/vectorsearch/milvus.rb', line 42

def remove_texts(ids:)
  raise ArgumentError, "ids must be an array" unless ids.is_a?(Array)

  client.entities.delete(
    collection_name: index_name,
    filter: "id in #{ids}"
  )
end

#similarity_search(query:, k: 4) ⇒ Object 



119
120
121
122
123
124
125
126
 
# File 'lib/langchain/vectorsearch/milvus.rb', line 119

def similarity_search(query:, k: 4)
  embedding = llm.embed(text: query).embedding

  similarity_search_by_vector(
    embedding: embedding,
    k: k
  )
end

#similarity_search_by_vector(embedding:, k: 4) ⇒ Object 



128
129
130
131
132
133
134
135
136
137
138
 
# File 'lib/langchain/vectorsearch/milvus.rb', line 128

def similarity_search_by_vector(embedding:, k: 4)
  load_default_schema

  client.entities.search(
    collection_name: index_name,
    anns_field: "vector",
    data: [embedding],
    limit: k,
    output_fields: ["content", "id", "vector"]
  )
end