Class: Ragnar::Embedder

Inherits:
Object
  • Object
show all
Defined in:
lib/ragnar/embedder.rb

Instance Attribute Summary collapse

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(model_name: Ragnar::DEFAULT_EMBEDDING_MODEL) ⇒ Embedder

Returns a new instance of Embedder.



5
6
7
8
 
# File 'lib/ragnar/embedder.rb', line 5

def initialize(model_name: Ragnar::DEFAULT_EMBEDDING_MODEL)
  @model_name = model_name
  @model = load_model(model_name)
end

Instance Attribute Details

#model ⇒ Object (readonly)

Returns the value of attribute model.



3
4
5
 
# File 'lib/ragnar/embedder.rb', line 3

def model
  @model
end

#model_name ⇒ Object (readonly)

Returns the value of attribute model_name.



3
4
5
 
# File 'lib/ragnar/embedder.rb', line 3

def model_name
  @model_name
end

Class Method Details

.available_models ⇒ Object 



89
90
91
92
93
94
95
96
97
98
99
100
101
102
 
# File 'lib/ragnar/embedder.rb', line 89

def self.available_models
  # List of commonly used embedding models
  # This could be expanded or made dynamic
  [
    "BAAI/bge-small-en-v1.5",
    "BAAI/bge-base-en-v1.5",
    "BAAI/bge-large-en-v1.5",
    "sentence-transformers/all-MiniLM-L6-v2",
    "sentence-transformers/all-mpnet-base-v2",
    "thenlper/gte-small",
    "thenlper/gte-base",
    "thenlper/gte-large"
  ]
end

.model_info(model_name) ⇒ Object 



104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
 
# File 'lib/ragnar/embedder.rb', line 104

def self.model_info(model_name)
  # Provide information about embedding models
  info = {
    "BAAI/bge-small-en-v1.5" => {
      dimensions: 384,
      max_tokens: 512,
      description: "Small, fast, good quality embeddings"
    },
    "BAAI/bge-base-en-v1.5" => {
      dimensions: 768,
      max_tokens: 512,
      description: "Balanced size and quality"
    },
    "BAAI/bge-large-en-v1.5" => {
      dimensions: 1024,
      max_tokens: 512,
      description: "Large, highest quality embeddings"
    },
    "sentence-transformers/all-MiniLM-L6-v2" => {
      dimensions: 384,
      max_tokens: 256,
      description: "Fast, lightweight model"
    },
    "sentence-transformers/all-mpnet-base-v2" => {
      dimensions: 768,
      max_tokens: 384,
      description: "High quality general purpose embeddings"
    }
  }
  
  info[model_name] || { description: "Model information not available" }
end

Instance Method Details

#embed_batch(texts, show_progress: true) ⇒ Object 



34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
 
# File 'lib/ragnar/embedder.rb', line 34

def embed_batch(texts, show_progress: true)
  embeddings = []
  
  if show_progress && $stdout.respond_to?(:ioctl)
    progressbar = TTY::ProgressBar.new(
      "Generating embeddings [:bar] :percent :current/:total",
      total: texts.size,
      bar_format: :block,
      width: 30
    )
  end
  
  texts.each do |text|
    embedding = embed_text(text)
    embeddings << embedding
    progressbar.advance if show_progress
  end
  
  embeddings
end

#embed_chunks(chunks, show_progress: true) ⇒ Object 



55
56
57
58
59
60
61
62
63
64
65
 
# File 'lib/ragnar/embedder.rb', line 55

def embed_chunks(chunks, show_progress: true)
  texts = chunks.map do |chunk|
    if chunk.is_a?(Hash)
      chunk[:text] || chunk["text"]
    else
      chunk.to_s
    end
  end
  
  embed_batch(texts, show_progress: show_progress)
end

#embed_text(text) ⇒ Object 



10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
 
# File 'lib/ragnar/embedder.rb', line 10

def embed_text(text)
  return nil if text.nil? || text.empty? || (text.respond_to?(:strip) && text.strip.empty?)
  
  # Use Candle to generate embeddings
  # The embedding method returns a tensor, we need to convert to array
  embedding = @model.embedding(text)
  
  # Convert tensor to array - Candle tensors need double to_a
  # First to_a gives [tensor], second to_a on the tensor gives the float array
  if embedding.respond_to?(:to_a)
    result = embedding.to_a
    if result.is_a?(Array) && result.first.respond_to?(:to_a)
      result.first.to_a
    else
      result
    end
  else
    embedding
  end
rescue => e
  puts "Error generating embedding: #{e.message}"
  nil
end