langchainrb 0.11.3 → 0.12.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: c678ae75bc25b0501223f5b6ffd396a9159af4d0ddd87ddb1657429ed2ba24ce
-  data.tar.gz: df50ef0a6d9c1a3100153a06084556cac983069d1a38739bd6606f39f63bd332
+  metadata.gz: 7f29aad35bc35dc95eb8673b11578b51c7449a19818989d9da5e640c6fb219c7
+  data.tar.gz: 4d0c4d3d424a82c7f02fb9e49ca52a5bdca5dfbce19fbfa22f2d74ef46d81eb7
 SHA512:
-  metadata.gz: 3ec9f92f4c6221184b7a0a2c118caa6a56e7bc8505a83d5b5acb4daeb769ff90d8822b43c28f57adc06435dd2df5577268721345c4061d3dad6ecb919be18efc
-  data.tar.gz: 53d54b0c6a82082438f2e2f1ca70d097a9b916bc283b72e52ce466b6f012c9624cf094586b17e93eaa49a796bf9911051d3f4b494b9ecc93c3ac6ee6cdc7e8fe
+  metadata.gz: 91b6f4fc5056308eab9119dcfda1be16857e6e9e6e531977148b1e8f31b72090794b67e6855afb95633b8f836b8d20921bc5a069afdc745d1114892143a177e1
+  data.tar.gz: f7a7949ab2efd960eacf3a93f7beaa9104403a93619b8c95ea094901c2d3d19b89980c81d293ae16035c5ff51fe021a09f2e81e2c0ed6854bff87d30e6def925

data/CHANGELOG.md

@@ -1,9 +1,13 @@
 ## [Unreleased]
-- New `Langchain::Processors::Pptx` to parse .pptx files
-- New `Langchain::LLM::Anthropic#chat()` support
-- Misc fixes
 
-## [0.11.3]
+## [0.12.0] - 2024-04-22
+- [BREAKING] Rename `dimension` parameter to `dimensions` everywhere
+
+## [0.11.4] - 2024-04-19
+- New `Langchain::LLM::AWSBedrock#chat()` to wrap Bedrock Claude requests
+- New `Langchain::LLM::OllamaResponse#total_tokens()` method
+
+## [0.11.3] - 2024-04-16
 - New `Langchain::Processors::Pptx` to parse .pptx files
 - New `Langchain::LLM::Anthropic#chat()` support
 - Misc fixes

data/README.md

@@ -60,7 +60,7 @@ Langchain.rb wraps supported LLMs in a unified interface allowing you to easily
 | [OpenAI](https://openai.com/?utm_source=langchainrb&utm_medium=github)                          | ✅                 | ✅                 | ✅                  | ❌                 | Including Azure OpenAI |
 | [AI21](https://ai21.com/?utm_source=langchainrb&utm_medium=github)                              | ❌                 | ✅                 | ❌                  | ✅                 |                    |
 | [Anthropic](https://anthropic.com/?utm_source=langchainrb&utm_medium=github)                    | ❌                 | ✅                 | ✅                  | ❌                 |                    |
-| [AWS Bedrock](https://aws.amazon.com/bedrock?utm_source=langchainrb&utm_medium=github)          | ✅                 | ✅                 | ❌                  | ❌                 | Provides AWS, Cohere, AI21, Antropic and Stability AI models |
+| [AWS Bedrock](https://aws.amazon.com/bedrock?utm_source=langchainrb&utm_medium=github)          | ✅                 | ✅                 | ✅                  | ❌                 | Provides AWS, Cohere, AI21, Antropic and Stability AI models |
 | [Cohere](https://cohere.com/?utm_source=langchainrb&utm_medium=github)                          | ✅                 | ✅                 | ✅                  | ✅                 |                    |
 | [GooglePalm](https://ai.google/discover/palm2?utm_source=langchainrb&utm_medium=github)         | ✅                 | ✅                 | ✅                  | ✅                 |                    |
 | [Google Vertex AI](https://cloud.google.com/vertex-ai?utm_source=langchainrb&utm_medium=github) | ✅                 | ✅                 | ❌                  | ✅                 |                    |

data/lib/langchain/llm/aws_bedrock.rb

@@ -46,7 +46,10 @@ module Langchain::LLM
       }
     }.freeze
 
+    attr_reader :client, :defaults
+
     SUPPORTED_COMPLETION_PROVIDERS = %i[anthropic cohere ai21].freeze
+    SUPPORTED_CHAT_COMPLETION_PROVIDERS = %i[anthropic].freeze
     SUPPORTED_EMBEDDING_PROVIDERS = %i[amazon].freeze
 
     def initialize(completion_model: DEFAULTS[:completion_model_name], embedding_model: DEFAULTS[:embedding_model_name], aws_client_options: {}, default_options: {})
@@ -91,6 +94,8 @@ module Langchain::LLM
     def complete(prompt:, **params)
       raise "Completion provider #{completion_provider} is not supported." unless SUPPORTED_COMPLETION_PROVIDERS.include?(completion_provider)
 
+      raise "Model #{@defaults[:completion_model_name]} only supports #chat." if @defaults[:completion_model_name].include?("claude-3")
+
       parameters = compose_parameters params
 
       parameters[:prompt] = wrap_prompt prompt
@@ -105,6 +110,53 @@ module Langchain::LLM
       parse_response response
     end
 
+    # Generate a chat completion for a given prompt
+    # Currently only configured to work with the Anthropic provider and
+    # the claude-3 model family
+    # @param messages [Array] The messages to generate a completion for
+    # @param system [String] The system prompt to provide instructions
+    # @param model [String] The model to use for completion defaults to @defaults[:chat_completion_model_name]
+    # @param max_tokens [Integer] The maximum number of tokens to generate
+    # @param stop_sequences [Array] The stop sequences to use for completion
+    # @param temperature [Float] The temperature to use for completion
+    # @param top_p [Float] The top p to use for completion
+    # @param top_k [Integer] The top k to use for completion
+    # @return [Langchain::LLM::AnthropicMessagesResponse] Response object
+    def chat(
+      messages: [],
+      system: nil,
+      model: defaults[:completion_model_name],
+      max_tokens: defaults[:max_tokens_to_sample],
+      stop_sequences: nil,
+      temperature: nil,
+      top_p: nil,
+      top_k: nil
+    )
+      raise ArgumentError.new("messages argument is required") if messages.empty?
+
+      raise "Model #{model} does not support chat completions." unless Langchain::LLM::AwsBedrock::SUPPORTED_CHAT_COMPLETION_PROVIDERS.include?(completion_provider)
+
+      inference_parameters = {
+        messages: messages,
+        max_tokens: max_tokens,
+        anthropic_version: @defaults[:anthropic_version]
+      }
+      inference_parameters[:system] = system if system
+      inference_parameters[:stop_sequences] = stop_sequences if stop_sequences
+      inference_parameters[:temperature] = temperature if temperature
+      inference_parameters[:top_p] = top_p if top_p
+      inference_parameters[:top_k] = top_k if top_k
+
+      response = client.invoke_model({
+        model_id: model,
+        body: inference_parameters.to_json,
+        content_type: "application/json",
+        accept: "application/json"
+      })
+
+      parse_response response
+    end
+
     private
 
     def completion_provider

data/lib/langchain/llm/base.rb

@@ -24,8 +24,8 @@ module Langchain::LLM
     # A client for communicating with the LLM
     attr_reader :client
 
-    def default_dimension
-      self.class.const_get(:DEFAULTS).dig(:dimension)
+    def default_dimensions
+      self.class.const_get(:DEFAULTS).dig(:dimensions)
     end
 
     #

data/lib/langchain/llm/cohere.rb

@@ -15,7 +15,7 @@ module Langchain::LLM
       temperature: 0.0,
       completion_model_name: "command",
       embeddings_model_name: "small",
-      dimension: 1024,
+      dimensions: 1024,
       truncate: "START"
     }.freeze
 

data/lib/langchain/llm/google_palm.rb

@@ -13,7 +13,7 @@ module Langchain::LLM
   class GooglePalm < Base
     DEFAULTS = {
       temperature: 0.0,
-      dimension: 768, # This is what the `embedding-gecko-001` model generates
+      dimensions: 768, # This is what the `embedding-gecko-001` model generates
       completion_model_name: "text-bison-001",
       chat_completion_model_name: "chat-bison-001",
       embeddings_model_name: "embedding-gecko-001"

data/lib/langchain/llm/google_vertex_ai.rb

@@ -16,7 +16,7 @@ module Langchain::LLM
       max_output_tokens: 1000,
       top_p: 0.8,
       top_k: 40,
-      dimension: 768,
+      dimensions: 768,
       completion_model_name: "text-bison", # Optional: tect-bison@001
       embeddings_model_name: "textembedding-gecko"
     }.freeze

data/lib/langchain/llm/hugging_face.rb

@@ -16,7 +16,7 @@ module Langchain::LLM
     DEFAULTS = {
       temperature: 0.0,
       embeddings_model_name: "sentence-transformers/all-MiniLM-L6-v2",
-      dimension: 384 # Vector size generated by the above model
+      dimensions: 384 # Vector size generated by the above model
     }.freeze
 
     #

data/lib/langchain/llm/ollama.rb

@@ -41,9 +41,9 @@ module Langchain::LLM
 
     # Returns the # of vector dimensions for the embeddings
     # @return [Integer] The # of vector dimensions
-    def default_dimension
+    def default_dimensions
       # since Ollama can run multiple models, look it up or generate an embedding and return the size
-      @default_dimension ||=
+      @default_dimensions ||=
         EMBEDDING_SIZES.fetch(defaults[:embeddings_model_name].to_sym) do
           embed(text: "test").embedding.size
         end

data/lib/langchain/llm/openai.rb

@@ -184,7 +184,7 @@ module Langchain::LLM
       complete(prompt: prompt)
     end
 
-    def default_dimension
+    def default_dimensions
       @defaults[:dimensions] || EMBEDDING_SIZES.fetch(defaults[:embeddings_model_name])
     end
 

data/lib/langchain/llm/replicate.rb

@@ -24,7 +24,7 @@ module Langchain::LLM
       # TODO: Design the interface to pass and use different models
       completion_model_name: "replicate/vicuna-13b",
       embeddings_model_name: "creatorrr/all-mpnet-base-v2",
-      dimension: 384
+      dimensions: 384
     }.freeze
 
     #

data/lib/langchain/llm/response/ollama_response.rb

@@ -44,5 +44,9 @@ module Langchain::LLM
     def completion_tokens
       raw_response.dig("eval_count")
     end
+
+    def total_tokens
+      prompt_tokens + completion_tokens
+    end
   end
 end

data/lib/langchain/vectorsearch/elasticsearch.rb

@@ -104,7 +104,7 @@ module Langchain::Vectorsearch
     end
 
     def default_vector_settings
-      {type: "dense_vector", dims: llm.default_dimension}
+      {type: "dense_vector", dims: llm.default_dimensions}
     end
 
     def vector_settings

data/lib/langchain/vectorsearch/epsilla.rb

@@ -54,7 +54,7 @@ module Langchain::Vectorsearch
       @db_path = db_path
       @table_name = index_name
 
-      @vector_dimension = llm.default_dimension
+      @vector_dimensions = llm.default_dimensions
 
       super(llm: llm)
     end
@@ -64,7 +64,7 @@ module Langchain::Vectorsearch
       status_code, response = @client.database.create_table(@table_name, [
         {"name" => "ID", "dataType" => "STRING", "primaryKey" => true},
         {"name" => "Doc", "dataType" => "STRING"},
-        {"name" => "Embedding", "dataType" => "VECTOR_FLOAT", "dimensions" => @vector_dimension}
+        {"name" => "Embedding", "dataType" => "VECTOR_FLOAT", "dimensions" => @vector_dimensions}
       ])
       raise "Failed to create table: #{response}" if status_code != 200
 

data/lib/langchain/vectorsearch/hnswlib.rb

@@ -26,7 +26,7 @@ module Langchain::Vectorsearch
 
       super(llm: llm)
 
-      @client = ::Hnswlib::HierarchicalNSW.new(space: DEFAULT_METRIC, dim: llm.default_dimension)
+      @client = ::Hnswlib::HierarchicalNSW.new(space: DEFAULT_METRIC, dim: llm.default_dimensions)
       @path_to_index = path_to_index
 
       initialize_index

data/lib/langchain/vectorsearch/milvus.rb

@@ -71,7 +71,7 @@ module Langchain::Vectorsearch
             type_params: [
               {
                 key: "dim",
-                value: llm.default_dimension.to_s
+                value: llm.default_dimensions.to_s
               }
             ]
           }

data/lib/langchain/vectorsearch/pgvector.rb

@@ -101,11 +101,11 @@ module Langchain::Vectorsearch
     def create_default_schema
       db.run "CREATE EXTENSION IF NOT EXISTS vector"
       namespace_column = @namespace_column
-      vector_dimension = llm.default_dimension
+      vector_dimensions = llm.default_dimensions
       db.create_table? table_name.to_sym do
         primary_key :id
         text :content
-        column :vectors, "vector(#{vector_dimension})"
+        column :vectors, "vector(#{vector_dimensions})"
         text namespace_column.to_sym, default: nil
       end
     end

data/lib/langchain/vectorsearch/pinecone.rb

@@ -104,7 +104,7 @@ module Langchain::Vectorsearch
       client.create_index(
         metric: DEFAULT_METRIC,
         name: index_name,
-        dimension: llm.default_dimension
+        dimension: llm.default_dimensions
       )
     end
 

data/lib/langchain/vectorsearch/qdrant.rb

@@ -93,7 +93,7 @@ module Langchain::Vectorsearch
         collection_name: index_name,
         vectors: {
           distance: DEFAULT_METRIC.capitalize,
-          size: llm.default_dimension
+          size: llm.default_dimensions
         }
       )
     end

data/lib/langchain/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Langchain
-  VERSION = "0.11.3"
+  VERSION = "0.12.0"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: langchainrb
 version: !ruby/object:Gem::Version
-  version: 0.11.3
+  version: 0.12.0
 platform: ruby
 authors:
 - Andrei Bondarev
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2024-04-17 00:00:00.000000000 Z
+date: 2024-04-22 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: activesupport