omniai 1.6.5 → 1.7.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 5d9011a55d57b9dd0923f4d7edd70001c3f9ee5b8211e98d59f9d001687af044
-  data.tar.gz: 204e2f0ec0a44f65cba11cb49819e7f5c4e933a92d6bc1b48c4d630dee180ecd
+  metadata.gz: 24e77ca1695f294daa0762f9c4c9b32a39ba329d8745d10d8b9fdd5c98917838
+  data.tar.gz: c6594df5e048bd4c7ea5d04bc3a66a01a2311ff2ff36ec5ee60db22afca7e412
 SHA512:
-  metadata.gz: 724827923255adade8f19bd9b43806d1af116d9c4dd6105713f0ba0b16c3d6cd5e5cbafdc2d324912a669523e565ba1ce190ed26035cdc1645a4505f2c6991a9
-  data.tar.gz: 5944ef3f285359662fc6105fd2a3aba7a44a279c5062ec0c1bdf194ddecd6842a33b1cd9b20cd4f5b36641b67cd8bc046e677260170d06dd2d5e3b37db37f356
+  metadata.gz: 274c978038080fbbb760dc3341a6ffdd103bd7ac176511189f1f36d8287084800dd5fdbb098138a3bb2b4405e7272763423b9815f1c2d1ca735a3f55f16271a4
+  data.tar.gz: 39e42f5e5ea5031653b2d6474326c767183270f6ca3dd1ff0b873c64fdc1999df1bcc58015457f5c8673b1c4544ef4d4fe39c72e7eb450951a839a5cc27b7382

data/README.md

@@ -229,6 +229,29 @@ tempfile.close
 tempfile.unlink
 ```
 
+### Embeddings
+
+Clients that support generating embeddings (e.g. OpenAI, Mistral, etc.) convert text to embeddings via the following:
+
+```ruby
+response = client.embed('The quick brown fox jumps over a lazy dog')
+response.usage # <OmniAI::Embed::Usage prompt_tokens=5 total_tokens=5>
+response.embedding # [0.1, 0.2, ...] >
+```
+
+Batches of text can also be converted to embeddings via the following:
+
+```ruby
+response = client.embed([
+  '',
+  '',
+])
+response.usage # <OmniAI::Embed::Usage prompt_tokens=5 total_tokens=5>
+response.embeddings.each do |embedding|
+  embedding # [0.1, 0.2, ...]
+end
+```
+
 ## CLI
 
 OmniAI packages a basic command line interface (CLI) to allow for exploration of various APIs. A detailed CLI documentation can be found via help:
@@ -263,3 +286,39 @@ Type 'exit' or 'quit' to abort.
 ```
 The warmest place on earth is Africa.
 ```
+
+### Embed
+
+#### w/ input
+
+```bash
+omniai embed "The quick brown fox jumps over a lazy dog."
+```
+
+```
+0.0
+...
+```
+
+#### w/o input
+
+```bash
+omniai embed --provider="openai" --model="text-embedding-ada-002"
+```
+
+```
+Type 'exit' or 'quit' to abort.
+# Whe quick brown fox jumps over a lazy dog.
+```
+
+```
+0.0
+...
+```
+
+0.0
+...
+
+```
+
+```

data/lib/omniai/chat/content.rb

@@ -4,6 +4,14 @@ module OmniAI
   class Chat
     # A placeholder for parts of a message. Any subclass must implement the serializable interface.
     class Content
+      # @return [String]
+      def self.summarize(content)
+        return content.map { |entry| summarize(entry) }.join("\n\n") if content.is_a?(Array)
+        return content if content.is_a?(String)
+
+        content.summarize
+      end
+
       # @param context [Context] optional
       #
       # @return [String]

data/lib/omniai/chat/media.rb

@@ -15,6 +15,11 @@ module OmniAI
         @type = type
       end
 
+      # @return [String]
+      def summarize
+        "[#{filename}]"
+      end
+
       # @return [Boolean]
       def text?
         @type.match?(%r{^text/})

data/lib/omniai/chat/message.rb

@@ -31,6 +31,14 @@ module OmniAI
         "#<#{self.class} role=#{@role.inspect} content=#{@content.inspect}>"
       end
 
+      # @return [String]
+      def summarize
+        <<~TEXT
+          #{@role}
+          #{Content.summarize(@content)}
+        TEXT
+      end
+
       # Usage:
       #
       #   Message.deserialize({ role: :user, content: 'Hello!' }) # => #<Message ...>

data/lib/omniai/chat/prompt.rb

@@ -62,6 +62,11 @@ module OmniAI
         "#<#{self.class.name} messages=#{@messages.inspect}>"
       end
 
+      # @return [String]
+      def summarize
+        @messages.map(&:summarize).join("\n\n")
+      end
+
       # Usage:
       #
       #   prompt.serialize # => [{ content: "What is the capital of Canada?", role: :user }]

data/lib/omniai/chat/text.rb

@@ -18,6 +18,11 @@ module OmniAI
         "#<#{self.class} text=#{@text.inspect}>"
       end
 
+      # @return [String]
+      def summarize
+        @text
+      end
+
       # @param data [Hash]
       def self.deserialize(data, context: nil)
         deserialize = context&.deserializers&.[](:text)

data/lib/omniai/chat/url.rb

@@ -16,6 +16,11 @@ module OmniAI
         @uri = uri
       end
 
+      # @return [String]
+      def summarize
+        "[#{filename}]"
+      end
+
       # @return [String]
       def inspect
         "#<#{self.class} uri=#{@uri.inspect}>"

data/lib/omniai/cli/embed_handler.rb

@@ -0,0 +1,58 @@
+# frozen_string_literal: true
+
+module OmniAI
+  class CLI
+    # Used for CLI usage of 'omnia embed'.
+    class EmbedHandler < BaseHandler
+      # @param argv [Array<String>]
+      def handle!(argv:)
+        parser.parse!(argv)
+
+        if argv.empty?
+          listen!
+        else
+          embed(input: argv.join(' '))
+        end
+      end
+
+      private
+
+      def listen!
+        @stdout.puts('Type "exit" or "quit" to leave.')
+
+        loop do
+          @stdout.print('# ')
+          @stdout.flush
+          input = @stdin.gets&.chomp
+
+          break if input.nil? || input.match?(/\A(exit|quit)\z/i)
+
+          embed(input:)
+        rescue Interrupt
+          break
+        end
+      end
+
+      # @param input [String]
+      def embed(input:)
+        response = client.embed(input, **@args)
+        @stdout.puts(response.embedding)
+      end
+
+      # @return [OptionParser]
+      def parser
+        OptionParser.new do |options|
+          options.banner = 'usage: omniai embed [options] "<prompt>"'
+
+          options.on('-h', '--help', 'help') do
+            @stdout.puts(options)
+            exit
+          end
+
+          options.on('-p', '--provider=PROVIDER', 'provider') { |provider| @provider = provider }
+          options.on('-m', '--model=MODEL', 'model') { |model| @args[:model] = model }
+        end
+      end
+    end
+  end
+end

data/lib/omniai/cli.rb

@@ -28,10 +28,14 @@ module OmniAI
       command = argv.shift
       return if command.nil?
 
-      case command
-      when 'chat' then ChatHandler.handle!(stdin: @stdin, stdout: @stdout, provider: @provider, argv:)
-      else raise Error, "unsupported command=#{command.inspect}"
-      end
+      handler =
+        case command
+        when 'chat' then ChatHandler
+        when 'embed' then EmbedHandler
+        else raise Error, "unsupported command=#{command.inspect}"
+        end
+
+      handler.handle!(stdin: @stdin, stdout: @stdout, provider: @provider, argv:)
     end
 
     private

data/lib/omniai/client.rb

@@ -173,5 +173,15 @@ module OmniAI
     def speak(input, model:, voice:, speed: nil, format: nil, &stream)
       raise NotImplementedError, "#{self.class.name}#speak undefined"
     end
+
+    # @raise [OmniAI::Error]
+    #
+    # @param input [String] required
+    # @param model [String] required
+    #
+    # @return [OmniAI::Embed::Embedding]
+    def embed(input, model:)
+      raise NotImplementedError, "#{self.class.name}#embed undefined"
+    end
   end
 end

data/lib/omniai/context.rb

@@ -0,0 +1,43 @@
+# frozen_string_literal: true
+
+module OmniAI
+  # Used to handle the setup of serializer / deserializer required per provide (e.g. Anthropic / Google / etc).
+  #
+  # Usage:
+  #
+  #   OmniAI::Context.build do |context|
+  #     context.serializers[:prompt] = (prompt, context:) -> { ... }
+  #     context.serializers[:message] = (prompt, context:) -> { ... }
+  #     context.serializers[:file] = (prompt, context:) -> { ... }
+  #     context.serializers[:text] = (prompt, context:) -> { ... }
+  #     context.serializers[:url] = (prompt, context:) -> { ... }
+  #     context.deserializers[:prompt] = (data, context:) -> { Prompt.new(...) }
+  #     context.deserializers[:message] = (data, context:) -> { Message.new(...) }
+  #     context.deserializers[:file] = (data, context:) -> { File.new(...) }
+  #     context.deserializers[:text] = (data, context:) -> { Text.new(...) }
+  #     context.deserializers[:url] = (data, context:) -> { URL.new(...) }
+  #   end
+  class Context
+    # @return [Hash]
+    attr_accessor :serializers
+
+    # @return [Hash]
+    attr_reader :deserializers
+
+    # @yield [context]
+    # @yieldparam context [Context]
+    #
+    # @return [Context]
+    def self.build(&block)
+      new.tap do |context|
+        block&.call(context)
+      end
+    end
+
+    # @return [Context]
+    def initialize
+      @serializers = {}
+      @deserializers = {}
+    end
+  end
+end

data/lib/omniai/embed/response.rb

@@ -0,0 +1,59 @@
+# frozen_string_literal: true
+
+module OmniAI
+  class Embed
+    # The response returned by the API.
+    class Response
+      # @return [Hash]
+      attr_accessor :data
+
+      # @param data [Hash]
+      # @param context [OmniAI::Context] optional
+      def initialize(data:, context: nil)
+        @data = data
+        @context = context
+      end
+
+      # @return [String]
+      def inspect
+        "#<#{self.class.name}>"
+      end
+
+      # @return [Usage]
+      def usage
+        @usage ||= begin
+          deserializer = @context&.deserializers&.[](:usage)
+
+          if deserializer
+            deserializer.call(@data, context: @context)
+          else
+            prompt_tokens = @data.dig('usage', 'prompt_tokens')
+            total_tokens = @data.dig('usage', 'total_tokens')
+
+            Usage.new(prompt_tokens:, total_tokens:)
+          end
+        end
+      end
+
+      # @param index [Integer] optional
+      #
+      # @return [Array<Float>]
+      def embedding(index: 0)
+        embeddings[index]
+      end
+
+      # @return [Array<Array<Float>>]
+      def embeddings
+        @embeddings ||= begin
+          deserializer = @context&.deserializers&.[](:embeddings)
+
+          if deserializer
+            deserializer.call(@data, context: @context)
+          else
+            @data['data'].map { |embedding| embedding['embedding'] }
+          end
+        end
+      end
+    end
+  end
+end

data/lib/omniai/embed/usage.rb

@@ -0,0 +1,26 @@
+# frozen_string_literal: true
+
+module OmniAI
+  class Embed
+    # Token usage returned by the API.
+    class Usage
+      # @return [Integer]
+      attr_accessor :prompt_tokens
+
+      # @return [Integer]
+      attr_accessor :total_tokens
+
+      # @param prompt_tokens Integer
+      # @param total_tokens Integer
+      def initialize(prompt_tokens:, total_tokens:)
+        @prompt_tokens = prompt_tokens
+        @total_tokens = total_tokens
+      end
+
+      # @return [String]
+      def inspect
+        "#<#{self.class.name} prompt_tokens=#{@prompt_tokens} total_tokens=#{@total_tokens}>"
+      end
+    end
+  end
+end

data/lib/omniai/embed.rb

@@ -0,0 +1,80 @@
+# frozen_string_literal: true
+
+module OmniAI
+  # An abstract class that provides a consistent interface for processing embedding requests.
+  #
+  # Usage:
+  #
+  #   class OmniAI::OpenAI::Embed < OmniAI::Embed
+  #     module Model
+  #       SMALL = "text-embedding-3-small"
+  #       LARGE = "text-embedding-3-large"
+  #       ADA = "text-embedding-3-002"
+  #     end
+  #
+  #     protected
+  #
+  #     # @return [Hash]
+  #     def payload
+  #       { ... }
+  #     end
+  #
+  #     # @return [String]
+  #     def path
+  #       "..."
+  #     end
+  #   end
+  #
+  #   client.embed(input, model: "...")
+  class Embed
+    def self.process!(...)
+      new(...).process!
+    end
+
+    # @param input [String] required
+    # @param client [Client] the client
+    # @param model [String] required
+    #
+    # @return [Response]
+    def initialize(input, client:, model:)
+      @input = input
+      @client = client
+      @model = model
+    end
+
+    # @raise [Error]
+    # @return [Response]
+    def process!
+      response = request!
+      raise HTTPError, response.flush unless response.status.ok?
+
+      parse!(response:)
+    end
+
+    protected
+
+    # @param response [HTTP::Response]
+    # @return [Response]
+    def parse!(response:)
+      Response.new(data: response.parse)
+    end
+
+    # @return [HTTP::Response]
+    def request!
+      @client
+        .connection
+        .accept(:json)
+        .post(path, json: payload)
+    end
+
+    # @return [Hash]
+    def payload
+      raise NotImplementedError, "#{self.class.name}#payload undefined"
+    end
+
+    # @return [String]
+    def path
+      raise NotImplementedError, "#{self.class.name}#path undefined"
+    end
+  end
+end

data/lib/omniai/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module OmniAI
-  VERSION = '1.6.5'
+  VERSION = '1.7.0'
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: omniai
 version: !ruby/object:Gem::Version
-  version: 1.6.5
+  version: 1.7.0
 platform: ruby
 authors:
 - Kevin Sylvestre
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2024-07-19 00:00:00.000000000 Z
+date: 2024-08-02 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: event_stream_parser
@@ -69,7 +69,6 @@ files:
 - lib/omniai.rb
 - lib/omniai/chat.rb
 - lib/omniai/chat/content.rb
-- lib/omniai/chat/context.rb
 - lib/omniai/chat/file.rb
 - lib/omniai/chat/media.rb
 - lib/omniai/chat/message.rb
@@ -93,8 +92,13 @@ files:
 - lib/omniai/cli.rb
 - lib/omniai/cli/base_handler.rb
 - lib/omniai/cli/chat_handler.rb
+- lib/omniai/cli/embed_handler.rb
 - lib/omniai/client.rb
 - lib/omniai/config.rb
+- lib/omniai/context.rb
+- lib/omniai/embed.rb
+- lib/omniai/embed/response.rb
+- lib/omniai/embed/usage.rb
 - lib/omniai/instrumentation.rb
 - lib/omniai/speak.rb
 - lib/omniai/tool.rb

data/lib/omniai/chat/context.rb

@@ -1,42 +0,0 @@
-# frozen_string_literal: true
-
-module OmniAI
-  class Chat
-    # Used to handle the setup of serializer / deserializer methods for each type.
-    #
-    # Usage:
-    #
-    #   OmniAI::Chat::Context.build do |context|
-    #     context.serializers[:prompt] = (prompt, context:) -> { ... }
-    #     context.serializers[:message] = (prompt, context:) -> { ... }
-    #     context.serializers[:file] = (prompt, context:) -> { ... }
-    #     context.serializers[:text] = (prompt, context:) -> { ... }
-    #     context.serializers[:url] = (prompt, context:) -> { ... }
-    #     context.deserializers[:prompt] = (data, context:) -> { Prompt.new(...) }
-    #     context.deserializers[:message] = (data, context:) -> { Message.new(...) }
-    #     context.deserializers[:file] = (data, context:) -> { File.new(...) }
-    #     context.deserializers[:text] = (data, context:) -> { Text.new(...) }
-    #     context.deserializers[:url] = (data, context:) -> { URL.new(...) }
-    #   end
-    class Context
-      # @return [Hash]
-      attr_accessor :serializers
-
-      # @return [Hash]
-      attr_reader :deserializers
-
-      # @return [Context]
-      def self.build(&block)
-        new.tap do |context|
-          block&.call(context)
-        end
-      end
-
-      # @return [Context]
-      def initialize
-        @serializers = {}
-        @deserializers = {}
-      end
-    end
-  end
-end