Class: Prescient::Provider::Ollama

Inherits:
Base
  • Object
show all
Includes:
HTTParty
Defined in:
lib/prescient/provider/ollama.rb

Overview

Ollama local or hosted API provider adapter.

Instance Attribute Summary

Attributes inherited from Base

#options, #provider_name

Instance Method Summary collapse

Methods inherited from Base

#available?, #build_prompt, #clean_text, #default_context_configs, #default_prompt_templates, #extract_embedding_text, #extract_text_values, #format_context_item, #handle_errors, #validate_embedding_dimensions

Constructor Details

#initialize(**options) ⇒ Ollama

Returns a new instance of Ollama.



9
10
11
12
13
 
# File 'lib/prescient/provider/ollama.rb', line 9

def initialize(**options)
  super
  self.class.base_uri(@options[:url])
  self.class.default_timeout(@options[:timeout] || 60)
end

Instance Method Details

#available_models ⇒ Array<Hash>

List models currently available from Ollama.

Returns:

  • (Array<Hash>) —

    Model descriptors including name, size, digest, modified_at, and booleans for configured embedding/chat roles



114
115
116
117
118
119
120
121
122
123
124
 
# File 'lib/prescient/provider/ollama.rb', line 114

def available_models
  return @_available_models if defined?(@_available_models)

  handle_errors do
    @_available_models = (fetch_and_parse('get', '/api/tags', root_key: 'models') || []).map { |model|
      { embedding:  model['name'] == @options[:embedding_model],
        chat:       model['name'] == @options[:chat_model],
        name: model['name'], size: model['size'], modified_at: model['modified_at'], digest: model['digest'] }
    }
  end
end

#generate_embedding(text, **_options) ⇒ Array<Float>

Generate an embedding through Ollama's /api/embed endpoint.

Parameters:

  • text (String) —

    Text to embed

Returns:

  • (Array<Float>) —

    Embedding vector from the first returned input item



18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
 
# File 'lib/prescient/provider/ollama.rb', line 18

def generate_embedding(text, **_options)
  handle_errors do
    embeddings = fetch_and_parse('post', '/api/embed',
                                 root_key: 'embeddings',
                                 headers:  { 'Content-Type' => 'application/json' },
                                 body:     {
                                   model: @options[:embedding_model],
                                   input: clean_text(text),
                                 }.to_json)

    embedding = embeddings.is_a?(Array) ? embeddings.first : nil # : Array[Float]?
    raise Prescient::InvalidResponseError, 'No embedding returned' unless embedding.is_a?(Array)

    expected_dimensions = @options[:embedding_dimensions]
    if expected_dimensions && embedding.length != expected_dimensions
      raise Prescient::InvalidResponseError,
            "Invalid embedding dimensions: expected #{expected_dimensions}, got #{embedding.length}"
    end

    embedding
  end
end

#generate_response(prompt, context_items = [], **options) ⇒ Hash

Generate text through Ollama's generation endpoint.

Parameters:

  • prompt (String) —

    Prompt to send

  • context_items (Array<Hash, String>) (defaults to: []) —

    Optional context items

Returns:

  • (Hash) —

    Normalized response data



45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
 
# File 'lib/prescient/provider/ollama.rb', line 45

def generate_response(prompt, context_items = [], **options)
  handle_errors do
    request_options = prepare_generate_response(prompt, context_items, **options)

    # Make the request and store both text and full response
    response = self.class.post('/api/generate', **request_options)
    validate_response!(response, 'POST /api/generate')

    generated_text = response.parsed_response['response']
    raise Prescient::InvalidResponseError, 'No response generated' unless generated_text

    {
      response:        generated_text.strip,
      model:           @options[:chat_model],
      provider:        'ollama',
      processing_time: response.parsed_response['total_duration']&./(1_000_000_000.0),
      metadata:        {
        eval_count:        response.parsed_response['eval_count'],
        eval_duration:     response.parsed_response['eval_duration'],
        prompt_eval_count: response.parsed_response['prompt_eval_count'],
      },
    }
  end
end

#health_check ⇒ Hash

Check whether the configured Ollama models are available locally.

reachable indicates the Ollama API answered successfully. ready indicates that both configured models are present in the local model list.

Returns:

  • (Hash) —

    Provider health information



76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
 
# File 'lib/prescient/provider/ollama.rb', line 76

def health_check
  handle_errors do
    models = available_models
    embedding_available = models.any? { |m| m[:embedding] }
    chat_available = models.any? { |m| m[:chat] }

    {
      status:           'healthy',
      provider:         'ollama',
      reachable:        true,
      url:              @options[:url],
      models_available: models.map { |m| m[:name] },
      embedding_model:  {
        name:      @options[:embedding_model],
        available: embedding_available,
      },
      chat_model:       {
        name:      @options[:chat_model],
        available: chat_available,
      },
      ready:            embedding_available && chat_available,
    }
  end
rescue Prescient::Error => e
  {
    status:    'unavailable',
    provider:  'ollama',
    reachable: false,
    error:     e.class.name,
    message:   e.message,
    url:       @options[:url],
    ready:     false,
  }
end

#pull_model(model_name) ⇒ Hash

Pull a model into the Ollama installation.

Parameters:

  • model_name (String) —

    Model identifier to download

Returns:

  • (Hash) —

    Pull result



129
130
131
132
133
134
135
136
137
138
139
140
141
 
# File 'lib/prescient/provider/ollama.rb', line 129

def pull_model(model_name)
  handle_errors do
    fetch_and_parse('post', '/api/pull',
                    headers: { 'Content-Type' => 'application/json' },
                    body:    { name: model_name }.to_json,
                    timeout: 300) # 5 minutes for model download
    {
      success: true,
      model:   model_name,
      message: "Model #{model_name} pulled successfully",
    }
  end
end

#validate_configuration! ⇒ Object (protected)

Raises:



145
146
147
148
149
150
151
152
 
# File 'lib/prescient/provider/ollama.rb', line 145

def validate_configuration!
  required_options = [:url, :embedding_model, :chat_model]
  missing_options = required_options.select { |opt| @options[opt].nil? }

  return unless missing_options.any?

  raise Prescient::Error, "Missing required options: #{missing_options.join(', ')}"
end