Usage - Azure OpenAI

Clojure functions to drive the Azure OpenAI API

Configuration

Add the openai-clojure dependency

deps.edn

net.clojars.wkok/openai-clojure {:mvn/version "0.18.1"}

Leiningen project.clj

[net.clojars.wkok/openai-clojure "0.18.1"]

Authentication

API Key

Set the environment variable AZURE_OPENAI_API_KEY to your Azure OpenAI API key.

An API key can be retrieved from your Azure account

Endpoint

Set the environment variable AZURE_OPENAI_API_ENDPOINT to your Azure OpenAPI endpoint, example: https://myendpoint.openai.azure.com

Options

Alternatively the api-key and/or api-endpoint can be passed in the options argument of each api function

(api/create-completion {:model "text-davinci-003"
                        :prompt "Say this is a test"}
                       {:api-key "xxxxx"
                        :api-endpoint "https://myendpoint.openai.azure.com"
                        :impl :azure})

Quickstart

See the full API Reference api documentation for examples of all the supported OpenAI APIs.

Require the api namespace

(:require [wkok.openai-clojure.api :as api])

A simple chat conversation with ChatGPT could be:

(api/create-chat-completion {:model "gpt-35-turbo"
                             :messages [{:role "system" :content "You are a helpful assistant."}
                                        {:role "user" :content "Who won the world series in 2020?"}
                                        {:role "assistant" :content "The Los Angeles Dodgers won the World Series in 2020."}
                                        {:role "user" :content "Where was it played?"}]}
                            {:impl :azure})

Result:

{:id "chatcmpl-6srOKLabYTpTRwRUQxjkcBxw3uf1H",
 :object "chat.completion",
 :created 1678532968,
 :model "gpt-35-turbo-0301",
 :usage {:prompt_tokens 56, :completion_tokens 19, :total_tokens 75},
 :choices
 [{:message
   {:role "assistant",
    :content
    "The 2020 World Series was played at Globe Life Field in Arlington, Texas."},
   :finish_reason "stop",
   :index 0}]}

Advanced configuration

Request options

Request options may be set on the underlying hato http client by adding a :request map to :options for example setting the request timeout:

(api/create-chat-completion {:model "gpt-3.5-turbo"
                             :messages [{:role "system" :content "You are a helpful assistant."}
                                        {:role "user" :content "Who won the world series in 2020?"}
                                        {:role "assistant" :content "The Los Angeles Dodgers won the World Series in 2020."}
                                        {:role "user" :content "Where was it played?"}]}

                            {:request {:timeout 60000}})    ;; <== This

Any of these supported request options may be passed.

Supported Azure OpenAI APIs

Completions

• create-completion

Also see the Azure OpenAI documentation

Chat

• create-chat-completion

Also see the Azure OpenAI documentation

Embeddings

• create-embedding

Also see the Azure OpenAI documentation