orbai

orbai: A Ruby DSL for OpenAI

Welcome to orbai, an idiomatic Ruby DSL for OpenAI's powerful API. With the combination of Ruby's natural language processing strengths and OpenAI's advanced capabilities, your applications are set to take a massive leap forward.

Disclaimer

Please be aware that orbai is in its early stages and is not maintained regularly. While we aim to provide the best, there might be bugs or unimplemented features. Your patience and contributions are highly appreciated.

Quick Start (For Rubyists)

1. Install the orbai gem:

gem install orbai

1. Have an OpenAI API key in your environment. By default, orbai will look for OPENAI_API_KEY..

2. Use orbai to query OpenAI:

require "orbai"

Orbai.query("What's the capital of France?")

# => "The capital of France is Paris."

Rationale

1. Ruby is the Best Language for DSLs: Ruby's expressiveness has made it the go-to language for Domain Specific Languages, finding its use in a myriad of practical domains.
2. OpenAI & Ruby: OpenAI offers transformative capabilities to the domains where Ruby DSLs excel.
3. orbai: Born out of the need for an expressive, easy-to-integrate tool that merges the power of OpenAI and the elegance of Ruby.

Getting Started (For Newbies)

1. Installing Ruby: If you're a technical person, we recommend using asdf-vm for version management.
2. Learning Ruby: If you're new to Ruby... TODO
3. Follow the "Quick Start" guide above to get up and running.

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.

Contributing

We welcome contributions to orbai! If you stumble upon issues or have features you'd like to suggest, please:

• Submit an issue
• For direct contributions, submit a pull request.

Help Wanted!

• 🔲 Feedback
• 🔲 Docs
• 🔲 Tests
• 🔲 Triage
• 🔲 Imagination

Orbai::DSL

Simple Query API

require "orbai"
Orbai.query("What is your favorite color?")
#=> "My favorite color is blue."

Wishlist

• 🔲 Config DX
• 🔲 Roles Builder DX
• 🔲 Context Params DX
• 🔲 Extract structured data
• 🔲 Functions DX
• 🔲 Images DX
• 🔲 Audio DX
• 🔲 Files DX
• 🔲 Embeddings DX
• 🔲 Fine-tuning DX

Orbai Module / HttpClient HTTP Methods

The Orbai module is just a container for everything else.

For ease of use, it includes http getters through an HttpClient singleton.

The HTTP getters throw an Orbai::HTTPClient::Error if the request fails. The OpenAI API type field from the error response is assigned to Orbai::HTTPClient::Error#type.

Example

Orbai.get("/chat/completions", model: 'gpt-3.5-turbo', messages: [
  {role:'system',content:'Act as a helpful prompt.'},
  {role:'user',content:'What is the capital of France?'},
])

# Returns
{"id"=>"chatcmpl-XXXXXXXX",
 "object"=>"chat.completion",
 "created"=>1697444538,
 "model"=>"gpt-3.5-turbo-0613",
 "choices"=>
  [{"index"=>0,
    "message"=>
     {"role"=>"assistant", "content"=>"The capital of France is Paris."},
    "finish_reason"=>"stop"}],
 "usage"=>{"prompt_tokens"=>24, "completion_tokens"=>7, "total_tokens"=>31}}

Orbai Shortcuts / HttpClient HTTP Methods

c = Orbai::HttpClient.new

# HttpClient singleton
c.http_client #=> #<Orbai::HttpClient>

# HTTP Methods
c.get(path, **params) #=> Returns JSON result as a ruby hash.
c.post(path, **params)
c.put(path, **params)
c.patch(path, **params)
c.delete(path, **params)
c.http_request(method, path, **params)

Orbai::HttpClient

The HttpClient class is used to make requests to the OpenAI API. It returns the JSON response from the API as a ruby hash.

Example:

c = Orbai::HttpClient.new

c.chat_completions(
  model: 'gpt-3.5-turbo',
  messages: [
    {role:'system',content:'Act as a helpful prompt.'},
    {role:'user',content:'What is the capital of France?'},
  ],
)

# Returns
{"id"=>"chatcmpl-XXXXXXXX",
 "object"=>"chat.completion",
 "created"=>1697983333,
 "model"=>"gpt-3.5-turbo-0613",
 "choices"=>
  [{"index"=>0,
    "message"=>
     {"role"=>"assistant", "content"=>"The capital of France is Paris."},
    "finish_reason"=>"stop"}],
 "usage"=>{"prompt_tokens"=>24, "completion_tokens"=>7, "total_tokens"=>31}}

API Methods Cheatsheet

c = Orbai::HttpClient.new

# Chat
#   POST /chat/completions
c.chat_completions(**params)

# Completions
#   POST /completions
c.completions(**params)

# Edits
#   POST /edits
c.edits(**params)

# Images
#   POST /images/generations
c.create_image(**params)
#   POST /images/edits
c.create_image_edit(**params)
#   POST /images/variations
c.create_image_variation(**params)

# Embeddings
#   POST /embeddings
c.create_embedding(**params)

# Fine_tuning
#
#   POST /fine_tuning/jobs
c.create_fine_tuning_jobs(**params)
#   GET /fine_tuning/jobs
c.list_fine_tuning_jobs(**params)
#   GET /fine_tuning/jobs/{fine_tuning_job_id}
c.retrieve_fine_tuning_job(id, **params)
#   POST /fine_tuning/jobs/{fine_tuning_job_id}/cancel
c.cancel_fine_tuning_job(id, **params)
#   GET /fine_tuning/jobs/{fine_tuning_job_id}/events
c.list_fine_tuning_events(id, **params)

# Audio
#   POST /audio/transcriptions
c.create_transcription(**params)
#   POST /audio/translations
c.create_translation(**params)

# Files
#   GET /files
c.list_files(**params)
#   POST /files
c.upload_file(**params)
#   DELETE /files/{file_id}
c.delete_file(id, **params)
#   GET /files/{file_id}
c.retrieve_file(id, **params)
#   GET /files/{file_id}/content
c.file_content(id, **params)

# Models
#   GET /models
c.list_models(**params)
#   GET /models/{model}
c.retrieve_model(id, **params)
#   DELETE /models/{model}
c.delete_model(id, **params)

# Moderations
#   POST /moderations
c.create_moderation(**params)