OpenaiChatgpt
OpenAI ChatGPT API is a light-weight Ruby wrapper for the Rubyists. It gives nice struct objects for completions of the chatgpt, even the raw responses are returned in nice open struct objects. It uses Faraday for HTTP requests.
It supports all models of gpt-4 and gpt-3.5
Installation
Install the gem and add to the application's Gemfile by executing:
$ bundle add openai_chatgpt
If bundler is not being used to manage dependencies, install the gem by executing:
$ gem install openai_chatgpt
Usage
API Documentation
The API documentation can be found at https://platform.openai.com/docs/api-reference/chat/create.
Get an API key
You can get API for the chatgpt by doing the following here.
Configuration
require 'openai_chatgpt'
client = OpenaiChatgpt::Client.new(api_key: ENV['CHATGPT_API_KEY'])
Params:
api_key
: String - Required - The API key for the openai chatgpt api.
Response:
#<OpenaiChatgpt::Client:0x00000001055fc318 @adapter=:net_http, @api_key="YOUR_API_KEY", @stubs=nil>
Get completion text
resp = OpenaiChatgpt::Client.new(api_key: "fake").completions(
messages: [
{role: "user", content: "Hello, how are you?"},
{role: "assistant", content: "I am good, how are you?"},
{role: "user", content: "What is your name?"},
]
)
Params:
- messages: Array[String] - Required - Messages to generate response for
[{role: "user", text: "Hello"}, {role: "bot", text: "Hi"}]
- model: String - Optional - Default: "gpt-4" - Model to use for generating response
["gpt-4", "gpt-4-0314", "gpt-4-32k", "gpt-4-32k-0314", "gpt-3.5-turbo", "gpt-3.5-turbo-0301"]
- temperature: Float - Optional - Default: 1.0 - Temperature for response generation between 0.0 and 2.0
- top_p: Float - Optional - Default: 1.0 - Top p for response generation between 0.0 and 1.0
- n: Integer - Optional - Default: 1 - Number of responses to generate
- stream: Boolean - Optional - Default: false - Stream response
- stop: Array[String] - Optional - Default: nil - Stop sequence for response generation
- max_tokens: Integer - Optional - Default: 4096 - Maximum number of tokens to generate
- presence_penalty: Float - Optional - Default: 0.0 - Presence penalty for response generation between -2.0 and 2.0
- frequency_penalty: Float - Optional - Default: 0.0 - Frequency penalty for response generation between
-2.0 and 2.0
- logit_bias: Hash - Optional - Default: nil - Logit bias for response generation
- user: String - Optional - Default: nil - User for response generation
Response:
resp.results
=> [#<OpenaiChatgpt::Choice:0x0000000107abca68 @content="I am an AI language model created by Open", @index=0, @role="assistant">]
resp.usage
=> #<OpenaiChatgpt::Usage:0x0000000107aba808 @completion_tokens=10, @prompt_tokens=36, @total_tokens=46>
Also raw_response
is available if you want the raw response from the API. This is useful if you want to do something with the raw response. Also that response is in OpenStruct Object which is a nice wrapper for the raw response and you can use 'dot' notation.
Error Handling
begin
client.completions(messages: [{role: "user", text: "Hello, how are you?"}])
rescue OpenaiChatgpt::Error => e
puts e.message
end
Development
After checking out the repo, run bin/setup
to install dependencies. Then, run rake test
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
Bug reports and pull requests are welcome on GitHub at https://github.com/sulmanweb/openai_chatgpt. This project is intended to be a safe, welcoming space for collaboration, and contributors are expected to adhere to the code of conduct.
License
The gem is available as open source under the terms of the MIT License.
Code of Conduct
Everyone interacting in the OpenaiChatgpt project's codebases, issue trackers, chat rooms and mailing lists is expected to follow the code of conduct.