llm_memory 0.1.0

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA256:
+  metadata.gz: c3f9540b8e3ca30d11eacd2e3100b4aa4ef985d041017b2b09bd5fad52d2f776
+  data.tar.gz: 9998bca0ea546d93b3a051ac7a6c62a9597ae2da9e85d8d5b36aab4212c59641
+SHA512:
+  metadata.gz: 9923ca576cf858eac89facf8da2f88d6284f7577366b252bb4e3ab197879a78035dfacba62e7890ab097dc193ee922f2b6abf16b321d4f1b09e6f1d30e4e9184
+  data.tar.gz: 44ff089feebead7cc9e1ef8b555212d20a938fcd6a4c72c13cfb55ae3029ebf6699722ea42e977ac108375e5d91c83aad450177ec2b9ad1037f87a83b83eadf7

data/.rspec

@@ -0,0 +1,3 @@
+--format documentation
+--color
+--require spec_helper

data/.standard.yml

@@ -0,0 +1,3 @@
+# For available configuration options, see:
+#   https://github.com/testdouble/standard
+ruby_version: 2.6

data/.vscode/settings.json

@@ -0,0 +1,7 @@
+{
+  "editor.formatOnSave": true,
+  "[ruby]": {
+    "editor.defaultFormatter": "testdouble.vscode-standard-ruby"
+  },
+  "standardRuby.autofix": true
+}

data/CHANGELOG.md

@@ -0,0 +1,5 @@
+## [Unreleased]
+
+## [0.1.0] - 2023-05-01
+
+- Initial release

data/CODE_OF_CONDUCT.md

@@ -0,0 +1,84 @@
+# Contributor Covenant Code of Conduct
+
+## Our Pledge
+
+We as members, contributors, and leaders pledge to make participation in our community a harassment-free experience for everyone, regardless of age, body size, visible or invisible disability, ethnicity, sex characteristics, gender identity and expression, level of experience, education, socio-economic status, nationality, personal appearance, race, religion, or sexual identity and orientation.
+
+We pledge to act and interact in ways that contribute to an open, welcoming, diverse, inclusive, and healthy community.
+
+## Our Standards
+
+Examples of behavior that contributes to a positive environment for our community include:
+
+* Demonstrating empathy and kindness toward other people
+* Being respectful of differing opinions, viewpoints, and experiences
+* Giving and gracefully accepting constructive feedback
+* Accepting responsibility and apologizing to those affected by our mistakes, and learning from the experience
+* Focusing on what is best not just for us as individuals, but for the overall community
+
+Examples of unacceptable behavior include:
+
+* The use of sexualized language or imagery, and sexual attention or
+  advances of any kind
+* Trolling, insulting or derogatory comments, and personal or political attacks
+* Public or private harassment
+* Publishing others' private information, such as a physical or email
+  address, without their explicit permission
+* Other conduct which could reasonably be considered inappropriate in a
+  professional setting
+
+## Enforcement Responsibilities
+
+Community leaders are responsible for clarifying and enforcing our standards of acceptable behavior and will take appropriate and fair corrective action in response to any behavior that they deem inappropriate, threatening, offensive, or harmful.
+
+Community leaders have the right and responsibility to remove, edit, or reject comments, commits, code, wiki edits, issues, and other contributions that are not aligned to this Code of Conduct, and will communicate reasons for moderation decisions when appropriate.
+
+## Scope
+
+This Code of Conduct applies within all community spaces, and also applies when an individual is officially representing the community in public spaces. Examples of representing our community include using an official e-mail address, posting via an official social media account, or acting as an appointed representative at an online or offline event.
+
+## Enforcement
+
+Instances of abusive, harassing, or otherwise unacceptable behavior may be reported to the community leaders responsible for enforcement at shoheik@cpan.org. All complaints will be reviewed and investigated promptly and fairly.
+
+All community leaders are obligated to respect the privacy and security of the reporter of any incident.
+
+## Enforcement Guidelines
+
+Community leaders will follow these Community Impact Guidelines in determining the consequences for any action they deem in violation of this Code of Conduct:
+
+### 1. Correction
+
+**Community Impact**: Use of inappropriate language or other behavior deemed unprofessional or unwelcome in the community.
+
+**Consequence**: A private, written warning from community leaders, providing clarity around the nature of the violation and an explanation of why the behavior was inappropriate. A public apology may be requested.
+
+### 2. Warning
+
+**Community Impact**: A violation through a single incident or series of actions.
+
+**Consequence**: A warning with consequences for continued behavior. No interaction with the people involved, including unsolicited interaction with those enforcing the Code of Conduct, for a specified period of time. This includes avoiding interactions in community spaces as well as external channels like social media. Violating these terms may lead to a temporary or permanent ban.
+
+### 3. Temporary Ban
+
+**Community Impact**: A serious violation of community standards, including sustained inappropriate behavior.
+
+**Consequence**: A temporary ban from any sort of interaction or public communication with the community for a specified period of time. No public or private interaction with the people involved, including unsolicited interaction with those enforcing the Code of Conduct, is allowed during this period. Violating these terms may lead to a permanent ban.
+
+### 4. Permanent Ban
+
+**Community Impact**: Demonstrating a pattern of violation of community standards, including sustained inappropriate behavior,  harassment of an individual, or aggression toward or disparagement of classes of individuals.
+
+**Consequence**: A permanent ban from any sort of public interaction within the community.
+
+## Attribution
+
+This Code of Conduct is adapted from the [Contributor Covenant][homepage], version 2.0,
+available at https://www.contributor-covenant.org/version/2/0/code_of_conduct.html.
+
+Community Impact Guidelines were inspired by [Mozilla's code of conduct enforcement ladder](https://github.com/mozilla/diversity).
+
+[homepage]: https://www.contributor-covenant.org
+
+For answers to common questions about this code of conduct, see the FAQ at
+https://www.contributor-covenant.org/faq. Translations are available at https://www.contributor-covenant.org/translations.

data/Gemfile

@@ -0,0 +1,15 @@
+# frozen_string_literal: true
+
+source "https://rubygems.org"
+
+# Specify your gem's dependencies in llm_memory.gemspec
+gemspec
+
+gem "rake", "~> 13.0"
+gem "rspec", "~> 3.0"
+gem "standard", "~> 1.3"
+gem "vcr", "~> 6.1.0"
+gem "webmock", "~> 3.18.1"
+gem "ruby-openai"
+gem "tiktoken_ruby"
+gem "redis"

data/Gemfile.lock

@@ -0,0 +1,107 @@
+PATH
+  remote: .
+  specs:
+    llm_memory (0.1.0)
+
+GEM
+  remote: https://rubygems.org/
+  specs:
+    addressable (2.8.4)
+      public_suffix (>= 2.0.2, < 6.0)
+    ast (2.4.2)
+    connection_pool (2.4.0)
+    crack (0.4.5)
+      rexml
+    diff-lcs (1.5.0)
+    faraday (2.7.4)
+      faraday-net_http (>= 2.0, < 3.1)
+      ruby2_keywords (>= 0.0.4)
+    faraday-multipart (1.0.4)
+      multipart-post (~> 2)
+    faraday-net_http (3.0.2)
+    hashdiff (1.0.1)
+    json (2.6.3)
+    language_server-protocol (3.17.0.3)
+    lint_roller (1.0.0)
+    multipart-post (2.3.0)
+    parallel (1.23.0)
+    parser (3.2.2.1)
+      ast (~> 2.4.1)
+    public_suffix (5.0.1)
+    rainbow (3.1.1)
+    rake (13.0.6)
+    redis (5.0.6)
+      redis-client (>= 0.9.0)
+    redis-client (0.14.1)
+      connection_pool
+    regexp_parser (2.8.0)
+    rexml (3.2.5)
+    rspec (3.12.0)
+      rspec-core (~> 3.12.0)
+      rspec-expectations (~> 3.12.0)
+      rspec-mocks (~> 3.12.0)
+    rspec-core (3.12.2)
+      rspec-support (~> 3.12.0)
+    rspec-expectations (3.12.3)
+      diff-lcs (>= 1.2.0, < 2.0)
+      rspec-support (~> 3.12.0)
+    rspec-mocks (3.12.5)
+      diff-lcs (>= 1.2.0, < 2.0)
+      rspec-support (~> 3.12.0)
+    rspec-support (3.12.0)
+    rubocop (1.50.2)
+      json (~> 2.3)
+      parallel (~> 1.10)
+      parser (>= 3.2.0.0)
+      rainbow (>= 2.2.2, < 4.0)
+      regexp_parser (>= 1.8, < 3.0)
+      rexml (>= 3.2.5, < 4.0)
+      rubocop-ast (>= 1.28.0, < 2.0)
+      ruby-progressbar (~> 1.7)
+      unicode-display_width (>= 2.4.0, < 3.0)
+    rubocop-ast (1.28.0)
+      parser (>= 3.2.1.0)
+    rubocop-performance (1.16.0)
+      rubocop (>= 1.7.0, < 2.0)
+      rubocop-ast (>= 0.4.0)
+    ruby-openai (4.0.0)
+      faraday (>= 1)
+      faraday-multipart (>= 1)
+    ruby-progressbar (1.13.0)
+    ruby2_keywords (0.0.5)
+    standard (1.28.0)
+      language_server-protocol (~> 3.17.0.2)
+      lint_roller (~> 1.0)
+      rubocop (~> 1.50.2)
+      standard-custom (~> 1.0.0)
+      standard-performance (~> 1.0.1)
+    standard-custom (1.0.0)
+      lint_roller (~> 1.0)
+    standard-performance (1.0.1)
+      lint_roller (~> 1.0)
+      rubocop-performance (~> 1.16.0)
+    tiktoken_ruby (0.0.4-arm64-darwin)
+    unicode-display_width (2.4.2)
+    vcr (6.1.0)
+    webmock (3.18.1)
+      addressable (>= 2.8.0)
+      crack (>= 0.3.2)
+      hashdiff (>= 0.4.0, < 2.0.0)
+
+PLATFORMS
+  arm64-darwin-22
+  x86_64-linux
+
+DEPENDENCIES
+  llm_memory!
+  rake (~> 13.0)
+  redis
+  rspec (~> 3.0)
+  ruby-openai
+  standard (~> 1.3)
+  tiktoken_ruby
+  vcr (~> 6.1.0)
+  webmock (~> 3.18.1)
+
+BUNDLED WITH
+   2.4.6

data/LICENSE.txt

@@ -0,0 +1,21 @@
+The MIT License (MIT)
+
+Copyright (c) 2023 Shohei Kameda
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in
+all copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+THE SOFTWARE.

data/README.md

@@ -0,0 +1,126 @@
+# 🧠 LLM Memory 🌊🐴
+
+LLM Memory is a Ruby gem designed to provide large language models (LLMs) like ChatGPT with memory using in-context learning.
+This enables better integration with systems such as Rails and web services while providing a more user-friendly and abstract interface based on brain terms.
+
+## Key Features
+
+- In-context learning through input prompt context insertion
+- Data connectors for various data sources
+- Inspired by the Python library, [LlamaIndex](https://github.com/jerryjliu/llama_index)
+- Focus on integration with existing systems
+- Easy-to-understand abstraction using brain-related terms
+- Plugin architecture for custom loader creation and extending LLM support
+
+## LLM Memory Components
+
+1. LlmMemory::Wernicke: Responsible for loading external data (currently from files). More loader types are planned for future development.
+
+> Wernicke's area in brain is involved in the comprehension of written and spoken language
+
+2. LlmMemory::Hippocampus: Handles interaction with vector databases to retrieve relevant information based on the query. Currently, Redis with the Redisearch module is used as the vector database. (Note that Redisearch is a proprietary modules and available on RedisCloud, which offers a free plan). The reason for choosing this is that Redis is commonly used and easy to integrate with web services.
+
+> Hippocampus in brain plays important roles in the consolidation of information from short-term memory to long-term memory
+
+3. LlmMemory::Broca: Responds to queries using memories provided by the Hippocampus component. ERB is used for prompt templates, and a variety of templates can be found online (e.g., in [LangChain Hub](https://github.com/hwchase17/langchain-hub#-prompts)).
+
+> Broca's area in brain is also known as the motor speech area.
+
+## Installation
+
+TODO: Replace `UPDATE_WITH_YOUR_GEM_NAME_PRIOR_TO_RELEASE_TO_RUBYGEMS_ORG` with your gem name right after releasing it to RubyGems.org. Please do not do it earlier due to security reasons. Alternatively, replace this section with instructions to install your gem from git if you don't plan to release to RubyGems.org.
+
+Install the gem and add to the application's Gemfile by executing:
+
+    $ bundle add UPDATE_WITH_YOUR_GEM_NAME_PRIOR_TO_RELEASE_TO_RUBYGEMS_ORG
+
+If bundler is not being used to manage dependencies, install the gem by executing:
+
+    $ gem install UPDATE_WITH_YOUR_GEM_NAME_PRIOR_TO_RELEASE_TO_RUBYGEMS_ORG
+
+### Setup
+
+Set environment variable `OPENAI_ACCESS_TOKEN` and `REDISCLOUD_URL`
+or set in initializer.
+
+```ruby
+LlmMemory.configure do |c|
+  c.openai_access_token = "xxxxx"
+  c.redis_url = "redis://xxxx:6379"
+end
+```
+
+## Usage
+
+To use LLM Memory, follow these steps:
+
+1. Install the gem: gem install llm_memory
+2. Set up Redis with Redisearch module enabled
+3. Configure LLM Memory to connect to your Redis instance
+4. Use LlmMemory::Wernicke to load data from your external sources
+5. Use LlmMemory::Hippocampus to search for relevant information based on user queries
+6. Create and use ERB templates with LlmMemory::Broca to generate responses based on the information retrieved
+
+```ruby
+docs = LlmMemory::Wernicke.load(:file, "/tmp/a_directory")
+# docs is just an array of hash.
+# You don't have to use load method but
+# create own hash with having content and metadata(optional)
+# docs = [{
+#   content: "Hi there",
+#   metadata: {
+#     file_name: "a.txt"
+#   }
+# },,,]
+
+hippocampus = LlmMemory::Hippocampus.new
+hippocampus.memorize(docs)
+
+query_str = "What is my name?"
+related_docs = hippocampus.query(query_str, limit: 3)
+#[{
+#   vector_score: "0.192698478699",
+#   content: "My name is Mike",
+#   metadata: { ... }
+#},,,]
+
+# ERB
+template = <<-TEMPLATE
+Context information is below.
+---------------------
+<% related_docs.each do |doc| %>
+<%= doc[:content] %>
+file: <%= doc[:metadata][:file_name] %>
+
+<% end %>
+---------------------
+Given the context information and not prior knowledge,
+answer the question: <%= query_str %>
+TEMPLATE
+
+broca = LlmMemory::Broca.new(prompt: tempate, model: 'gpt-3.5-turbo')
+messages = broca.respond(query_str: query_str, related_docs: related_docs)
+
+...
+query_str2 = "How are you?"
+related_docs = hippocampus.query(query_str2, limit: 3)
+message2 = broca.respond(query_str: query_str2, related_docs: related_docs)
+```
+
+## Development
+
+After checking out the repo, run `bin/setup` to install dependencies. Then, run `rake spec` to run the tests. You can also run `bin/console` for an interactive prompt that will allow you to experiment.
+
+To install this gem onto your local machine, run `bundle exec rake install`. To release a new version, update the version number in `version.rb`, and then run `bundle exec rake release`, which will create a git tag for the version, push git commits and the created tag, and push the `.gem` file to [rubygems.org](https://rubygems.org).
+
+## Contributing
+
+Bug reports and pull requests are welcome on GitHub at https://github.com/[USERNAME]/llm_memory. This project is intended to be a safe, welcoming space for collaboration, and contributors are expected to adhere to the [code of conduct](https://github.com/[USERNAME]/llm_memory/blob/master/CODE_OF_CONDUCT.md).
+
+## License
+
+The gem is available as open source under the terms of the [MIT License](https://opensource.org/licenses/MIT).
+
+## Code of Conduct
+
+Everyone interacting in the LlmMemory project's codebases, issue trackers, chat rooms and mailing lists is expected to follow the [code of conduct](https://github.com/[USERNAME]/llm_memory/blob/master/CODE_OF_CONDUCT.md).

data/Rakefile

@@ -0,0 +1,10 @@
+# frozen_string_literal: true
+
+require "bundler/gem_tasks"
+require "rspec/core/rake_task"
+
+RSpec::Core::RakeTask.new(:spec)
+
+require "standard/rake"
+
+task default: %i[spec standard]

data/lib/llm_memory/broca.rb

@@ -0,0 +1,63 @@
+require "erb"
+require "tiktoken_ruby"
+
+module LlmMemory
+  class Broca
+    include Llms::Openai
+    attr_accessor :messages
+
+    def initialize(
+      prompt:,
+      model: "gpt-3.5-turbo",
+      temperature: 0.7,
+      max_token: 4096
+    )
+      @prompt = prompt
+      @model = model
+      @messages = []
+      @temperature = temperature
+      @max_token = max_token
+    end
+
+    def respond(*args)
+      final_prompt = generate_prompt(*args)
+      @messages.push({role: "user", content: final_prompt})
+      adjust_token_count
+      response = client.chat(
+        parameters: {
+          model: @model,
+          messages: @messages,
+          temperature: @temperature
+        }
+      )
+      response_conent = response.dig("choices", 0, "message", "content")
+      @messages.push({role: "system", content: response_conent})
+      response_conent
+    end
+
+    def generate_prompt(*args)
+      merged_args = args.reduce(:merge)
+      erb = ERB.new(@prompt)
+      erb.result_with_hash(merged_args)
+    end
+
+    def adjust_token_count
+      count = 0
+      new_messages = []
+      @messages.reverse_each do |message|
+        encoded = tokenizer.encode(message[:content])
+        if count < @max_token
+          count += encoded.length
+          new_messages.push(message)
+        else
+          break
+        end
+      end
+      @messages = new_messages.reverse
+    end
+
+    def tokenizer
+      @tokenizer ||= Tiktoken.encoding_for_model("gpt-4")
+    end
+  end
+end

data/lib/llm_memory/configuration.rb

@@ -0,0 +1,11 @@
+module LlmMemory
+  class Configuration
+    attr_accessor :openai_access_token, :openai_organization_id, :redis_url
+
+    def initialize
+      @openai_access_token = ENV["OPENAI_ACCESS_TOKEN"]
+      @openai_organization_id = nil
+      @redis_url = ENV["REDISCLOUD_URL"] || "redis://localhost:6379"
+    end
+  end
+end

data/lib/llm_memory/embedding.rb

@@ -0,0 +1,30 @@
+# lib/lilyama_index/Embedding.rb
+module LlmMemory
+  module Embedding
+    def self.included(base)
+      base.extend(ClassMethods)
+    end
+
+    module ClassMethods
+      def register_embedding(name)
+        LlmMemory::EmbeddingManager.register_embedding(name, self)
+      end
+    end
+
+    def embed_document(text)
+      raise NotImplementedError, "Each Embedding must implement the 'embed_document' method."
+    end
+  end
+
+  class EmbeddingManager
+    @embeddings = {}
+
+    def self.register_embedding(name, klass)
+      @embeddings[name] = klass
+    end
+
+    def self.embeddings
+      @embeddings
+    end
+  end
+end

data/lib/llm_memory/embeddings/openai.rb

@@ -0,0 +1,37 @@
+require_relative "../embedding"
+require_relative "../llms/openai"
+
+module LlmMemory
+  module Embeddings
+    class Openai
+      include LlmMemory::Embedding
+      include Llms::Openai
+
+      register_embedding :openai
+
+      def embed_documents(texts, model: "text-embedding-ada-002")
+        embedding_list = []
+        texts.each do |txt|
+          res = client.embeddings(
+            parameters: {
+              model: model,
+              input: txt
+            }
+          )
+          embedding_list.push(res["data"][0]["embedding"])
+        end
+        embedding_list
+      end
+
+      def embed_document(text, model: "text-embedding-ada-002")
+        res = client.embeddings(
+          parameters: {
+            model: model,
+            input: text
+          }
+        )
+        res["data"][0]["embedding"]
+      end
+    end
+  end
+end

data/lib/llm_memory/hippocampus.rb

@@ -0,0 +1,95 @@
+require_relative "store"
+require_relative "stores/redis_store"
+
+require_relative "embedding"
+require_relative "embeddings/openai"
+
+module LlmMemory
+  class Hippocampus
+    def initialize(
+      embedding_name: :openai,
+      chunk_size: 1024,
+      chunk_overlap: 50,
+      store_name: :redis,
+      index_name: "llm_memory"
+    )
+      embedding_class = EmbeddingManager.embeddings[embedding_name]
+      raise "Embedding '#{embedding_name}' not found." unless embedding_class
+      @embedding_instance = embedding_class.new
+
+      store_class = StoreManager.stores[store_name]
+      raise "Store '#{store_name}' not found." unless store_class
+      @store = store_class.new(index_name: index_name)
+
+      # word count, not char count
+      @chunk_size = chunk_size
+      @chunk_overlap = chunk_overlap
+    end
+
+    def memorize(docs)
+      docs = make_chunks(docs)
+      docs = add_vectors(docs)
+      @store.create_index unless @store.index_exists?
+      @store.add(data: docs)
+    end
+
+    def query(query_str, limit: 3)
+      vector = @embedding_instance.embed_document(query_str)
+      response_list = @store.search(query: vector, k: limit)
+      response_list.shift # the first one is the size
+      # now [redis_key1, [],,, ]
+      result = response_list.each_slice(2).to_h.values.map { |v|
+        v.each_slice(2).to_h.transform_keys(&:to_sym)
+      }
+      result.each do |item|
+        item[:metadata] = JSON.parse(item[:metadata])
+      end
+      result
+    end
+
+    def forgot_all
+      @store.drop_index
+    end
+
+    def add_vectors(docs)
+      # embed documents and add vector
+      result = []
+      docs.each do |doc|
+        content = doc[:content]
+        metadata = doc[:metadata]
+        vector = @embedding_instance.embed_document(content)
+        result.push({
+          content: content,
+          metadata: metadata,
+          vector: vector
+        })
+      end
+      result
+    end
+
+    def make_chunks(docs)
+      result = []
+      docs.each do |item|
+        content = item[:content]
+        metadata = item[:metadata]
+        words = content.split
+
+        if words.length > @chunk_size
+          start_index = 0
+
+          while start_index < words.length
+            end_index = [start_index + @chunk_size, words.length].min
+            chunk_words = words[start_index...end_index]
+            chunk = chunk_words.join(" ")
+            result << {content: chunk, metadata: metadata}
+
+            start_index += @chunk_size - @chunk_overlap # Move index to create a overlap
+          end
+        else
+          result << {content: content, metadata: metadata}
+        end
+      end
+      result
+    end
+  end
+end

data/lib/llm_memory/llms/openai.rb

@@ -0,0 +1,14 @@
+require "openai"
+require "llm_memory"
+
+module LlmMemory
+  module Llms
+    module Openai
+      def client
+        @client ||= OpenAI::Client.new(
+          access_token: LlmMemory.configuration.openai_access_token
+        )
+      end
+    end
+  end
+end

data/lib/llm_memory/loader.rb

@@ -0,0 +1,30 @@
+# lib/lilyama_index/loader.rb
+module LlmMemory
+  module Loader
+    def self.included(base)
+      base.extend(ClassMethods)
+    end
+
+    module ClassMethods
+      def register_loader(name)
+        LlmMemory::LoaderManager.register_loader(name, self)
+      end
+    end
+
+    def load
+      raise NotImplementedError, "Each loader must implement the 'load' method."
+    end
+  end
+
+  class LoaderManager
+    @loaders = {}
+
+    def self.register_loader(name, klass)
+      @loaders[name] = klass
+    end
+
+    def self.loaders
+      @loaders
+    end
+  end
+end

data/lib/llm_memory/loaders/file_loader.rb

@@ -0,0 +1,29 @@
+require "find"
+require_relative "../loader"
+
+module LlmMemory
+  class FileLoader
+    include Loader
+
+    register_loader :file
+
+    def load(directory_path)
+      files_array = []
+      Find.find(directory_path) do |file_path|
+        next if File.directory?(file_path)
+
+        file_name = File.basename(file_path)
+        file_content = File.read(file_path)
+
+        files_array << {
+          content: file_content,
+          metadata: {
+            file_name: file_name
+          }
+        }
+      end
+
+      files_array
+    end
+  end
+end

data/lib/llm_memory/store.rb

@@ -0,0 +1,42 @@
+# lib/lilyama_index/store.rb
+module LlmMemory
+  module Store
+    def self.included(base)
+      base.extend(ClassMethods)
+    end
+
+    module ClassMethods
+      def register_store(name)
+        LlmMemory::StoreManager.register_store(name, self)
+      end
+    end
+
+    def create_index
+      raise NotImplementedError, "Each store must implement the 'create_index' method."
+    end
+
+    def drop_index
+      raise NotImplementedError, "Each store must implement the 'drop_index' method."
+    end
+
+    def add
+      raise NotImplementedError, "Each store must implement the 'add' method."
+    end
+
+    def search
+      raise NotImplementedError, "Each store must implement the 'search' method."
+    end
+  end
+
+  class StoreManager
+    @stores = {}
+
+    def self.register_store(name, klass)
+      @stores[name] = klass
+    end
+
+    def self.stores
+      @stores
+    end
+  end
+end

data/lib/llm_memory/stores/redis_store.rb

@@ -0,0 +1,154 @@
+require "redis"
+require_relative "../store"
+require "json"
+
+module LlmMemory
+  class RedisStore
+    include Store
+
+    register_store :redis
+
+    def initialize(
+      index_name: "llm_memory",
+      content_key: "content",
+      vector_key: "vector",
+      metadata_key: "metadata"
+    )
+      @index_name = index_name
+      @content_key = content_key
+      @vector_key = vector_key
+      @metadata_key = metadata_key
+      @client = Redis.new(url: LlmMemory.configuration.redis_url)
+    end
+
+    def info
+      @client.call(["INFO"])
+    end
+
+    def load_data(file_path)
+      CSV.read(file_path)
+    end
+
+    def list_indexes
+      @client.call("FT._LIST")
+    end
+
+    def index_exists?
+      begin
+        @client.call(["FT.INFO", @index_name])
+      rescue
+        return false
+      end
+      true
+    end
+
+    def drop_index
+      # DD deletes all document hashes
+      @client.call(["FT.DROPINDEX", @index_name, "DD"])
+    end
+
+    # dimention: 1536 for ada-002
+    def create_index(dim: 1536, distance_metric: "COSINE")
+      # LangChain index
+      # schema = (
+      #   TextField(name=content_key),
+      #   TextField(name=metadata_key),
+      #   VectorField(
+      #       vector_key,
+      #       "FLAT",
+      #       {
+      #           "TYPE": "FLOAT32",
+      #           "DIM": dim,
+      #           "DISTANCE_METRIC": distance_metric,
+      #       },
+      #   ),
+      # )
+      command = [
+        "FT.CREATE", @index_name, "ON", "HASH",
+        "PREFIX", "1", "#{@index_name}:",
+        "SCHEMA",
+        @content_key, "TEXT",
+        @metadata_key, "TEXT",
+        @vector_key, "VECTOR", "FLAT", 6, "TYPE", "FLOAT32", "DIM", dim, "DISTANCE_METRIC", distance_metric
+      ]
+      @client.call(command)
+    end
+
+    # data = [{ content: "", content_vector: [], metadata: {} }]
+    def add(data: [])
+      result = {}
+      @client.pipelined do |pipeline|
+        data.each_with_index do |d, i|
+          key = "#{@index_name}:#{SecureRandom.uuid.delete("-")}"
+          meta_json = d[:metadata].nil? ? "" : d[:metadata].to_json # serialize
+          vector_value = d[:vector].map(&:to_f).pack("f*")
+          pipeline.hset(
+            key,
+            {
+              @content_key => d[:content],
+              @vector_key => vector_value,
+              @metadata_key => meta_json
+            }
+          )
+          result[key] = d[:content]
+        end
+      end
+      result
+    # data.each_with_index do |d, i|
+    #   key = "#{@index_name}:#{i}"
+    #   vector_value = d[:content_vector].map(&:to_f).pack("f*")
+    #   pp vector_value
+    #   @client.hset(
+    #     key,
+    #     {
+    #       @content_key => d[:content],
+    #       @vector_key => vector_value,
+    #       @metadata_key => ""
+    #     }
+    #   )
+    # end
+    # rescue Redis::Pipeline::Error => e
+    #   # Handle the error if there is any issue with the pipeline execution
+    #   puts "Pipeline Error: #{e.message}"
+    # rescue Redis::BaseConnectionError => e
+    #   # Handle connection errors
+    #   puts "Connection Error: #{e.message}"
+    rescue => e
+      # Handle any other errors
+      puts "Unexpected Error: #{e.message}"
+    end
+
+    def delete
+    end
+
+    def update
+    end
+
+    def search(query: [], k: 3)
+      packed_query = query.map(&:to_f).pack("f*")
+      command = [
+        "FT.SEARCH",
+        @index_name,
+        "*=>[KNN #{k} @vector $blob AS vector_score]",
+        "PARAMS",
+        2,
+        "blob",
+        packed_query,
+        "SORTBY",
+        "vector_score",
+        "ASC",
+        "LIMIT",
+        0,
+        k,
+        "RETURN",
+        3,
+        "vector_score",
+        @content_key,
+        @metadata_key,
+        "DIALECT",
+        2
+      ]
+      @client.call(command)
+    end
+  end
+end

data/lib/llm_memory/version.rb

@@ -0,0 +1,5 @@
+# frozen_string_literal: true
+
+module LlmMemory
+  VERSION = "0.1.0"
+end

data/lib/llm_memory/wernicke.rb

@@ -0,0 +1,14 @@
+# loader
+require_relative "loader"
+require_relative "loaders/file_loader"
+
+module LlmMemory
+  class Wernicke
+    def self.load(loader_name, *args)
+      loader_class = LoaderManager.loaders[loader_name]
+      raise "Loader '#{loader_name}' not found." unless loader_class
+      loader_instance = loader_class.new
+      loader_instance.load(*args)
+    end
+  end
+end

data/lib/llm_memory.rb

@@ -0,0 +1,24 @@
+# frozen_string_literal: true
+
+# config
+require_relative "llm_memory/configuration"
+
+require_relative "llm_memory/hippocampus"
+require_relative "llm_memory/broca"
+require_relative "llm_memory/wernicke"
+
+require_relative "llm_memory/version"
+
+module LlmMemory
+  class Error < StandardError; end
+
+  class << self
+    attr_accessor :configuration
+  end
+
+  def self.configure
+    self.configuration ||= Configuration.new
+    yield(configuration) if block_given?
+  end
+  configure # init for default values
+end

data/llm_memory.gemspec

@@ -0,0 +1,37 @@
+# frozen_string_literal: true
+
+require_relative "lib/llm_memory/version"
+
+Gem::Specification.new do |spec|
+  spec.name = "llm_memory"
+  spec.version = LlmMemory::VERSION
+  spec.authors = ["Shohei Kameda"]
+  spec.email = ["shoheik@cpan.org"]
+
+  spec.summary = "A Ruby Gem for LLMs like ChatGPT to have memory using in-context learning"
+  spec.description = "LLM Memory is a Ruby gem designed to provide large language models (LLMs) like ChatGPT with memory using in-context learning. This enables better integration with systems such as Rails and web services while providing a more user-friendly and abstract interface based on brain terms."
+  spec.homepage = "https://github.com/shohey1226/llm_memory"
+  spec.license = "MIT"
+  spec.required_ruby_version = ">= 2.6.0"
+
+  spec.metadata["homepage_uri"] = spec.homepage
+  spec.metadata["source_code_uri"] = "https://github.com/shohey1226/llm_memory"
+  spec.metadata["changelog_uri"] = "https://github.com/shohey1226/llm_memory/CHANGELOG.md"
+
+  # Specify which files should be added to the gem when it is released.
+  # The `git ls-files -z` loads the files in the RubyGem that have been added into git.
+  spec.files = Dir.chdir(__dir__) do
+    `git ls-files -z`.split("\x0").reject do |f|
+      (f == __FILE__) || f.match(%r{\A(?:(?:bin|test|spec|features)/|\.(?:git|circleci)|appveyor)})
+    end
+  end
+  spec.bindir = "exe"
+  spec.executables = spec.files.grep(%r{\Aexe/}) { |f| File.basename(f) }
+  spec.require_paths = ["lib"]
+
+  # Uncomment to register a new dependency of your gem
+  # spec.add_dependency "example-gem", "~> 1.0"
+
+  # For more information and examples about making a new gem, check out our
+  # guide at: https://bundler.io/guides/creating_gem.html
+end

data/sig/llm_memory.rbs

@@ -0,0 +1,4 @@
+module LlmMemory
+  VERSION: String
+  # See the writing guide of rbs: https://github.com/ruby/rbs#guides
+end

metadata

@@ -0,0 +1,74 @@
+--- !ruby/object:Gem::Specification
+name: llm_memory
+version: !ruby/object:Gem::Version
+  version: 0.1.0
+platform: ruby
+authors:
+- Shohei Kameda
+autorequire:
+bindir: exe
+cert_chain: []
+date: 2023-05-04 00:00:00.000000000 Z
+dependencies: []
+description: LLM Memory is a Ruby gem designed to provide large language models (LLMs)
+  like ChatGPT with memory using in-context learning. This enables better integration
+  with systems such as Rails and web services while providing a more user-friendly
+  and abstract interface based on brain terms.
+email:
+- shoheik@cpan.org
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- ".rspec"
+- ".standard.yml"
+- ".vscode/settings.json"
+- CHANGELOG.md
+- CODE_OF_CONDUCT.md
+- Gemfile
+- Gemfile.lock
+- LICENSE.txt
+- README.md
+- Rakefile
+- lib/llm_memory.rb
+- lib/llm_memory/broca.rb
+- lib/llm_memory/configuration.rb
+- lib/llm_memory/embedding.rb
+- lib/llm_memory/embeddings/openai.rb
+- lib/llm_memory/hippocampus.rb
+- lib/llm_memory/llms/openai.rb
+- lib/llm_memory/loader.rb
+- lib/llm_memory/loaders/file_loader.rb
+- lib/llm_memory/store.rb
+- lib/llm_memory/stores/redis_store.rb
+- lib/llm_memory/version.rb
+- lib/llm_memory/wernicke.rb
+- llm_memory.gemspec
+- sig/llm_memory.rbs
+homepage: https://github.com/shohey1226/llm_memory
+licenses:
+- MIT
+metadata:
+  homepage_uri: https://github.com/shohey1226/llm_memory
+  source_code_uri: https://github.com/shohey1226/llm_memory
+  changelog_uri: https://github.com/shohey1226/llm_memory/CHANGELOG.md
+post_install_message:
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: 2.6.0
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubygems_version: 3.4.6
+signing_key:
+specification_version: 4
+summary: A Ruby Gem for LLMs like ChatGPT to have memory using in-context learning
+test_files: []