Terraform for setting up OpenAI APIs on Microsoft Azure

main.tf

# <[email protected]>, 2023-06-02
#
# Terraform for setting up GPT-4, GPT-3.5-turbo, and text-embedding-ada-002
# endpoints. Note that not all models are available in all regions, so 
# check before changing the region here, currently set to "southcentralus"
#
# Assumes az-cli authenticated (requires Azure subscription) and terraform 
# available and installed
#
terraform {
  required_providers {
    azurerm = {
      source  = "hashicorp/azurerm"
    }
  }
}

# Configure the Microsoft Azure Provider
provider "azurerm" {
  features {}
}

resource "azurerm_resource_group" "openai" {
  name     = "openai-rg"
  location = "eastus"
}

resource "azurerm_cognitive_account" "openai" {
  name                = "openai-account"
  location            = azurerm_resource_group.openai.location
  resource_group_name = azurerm_resource_group.openai.name
  kind                = "OpenAI"
  sku_name            = "S0"
}

resource "azurerm_cognitive_deployment" "embedding" {
  name                 = "embedding"
  cognitive_account_id = azurerm_cognitive_account.openai.id
  model {
    format = "OpenAI"
    name    = "text-embedding-ada-002"
    version = "2"
  }

  scale {
    type = "Standard"
  }
}

resource "azurerm_cognitive_deployment" "gpt_35_turbo" {
  name                 = "gpt-35-turbo"
  cognitive_account_id = azurerm_cognitive_account.openai.id
  model {
    format = "OpenAI"
    name    = "gpt-35-turbo"
    version = "0301"
  }

  scale {
    type = "Standard"
  }
}

resource "azurerm_cognitive_deployment" "gpt_4" {
  name                 = "gpt-4"
  cognitive_account_id = azurerm_cognitive_account.openai.id
  model {
    format = "OpenAI"
    name    = "gpt-4"
    version = "0314"
  }

  scale {
    type = "Standard"
  }
}

resource "azurerm_cognitive_deployment" "gpt_4_32k" {
  name                 = "gpt-4-32k"
  cognitive_account_id = azurerm_cognitive_account.openai.id
  model {
    format = "OpenAI"
    name    = "gpt-4-32k"
    version = "0314"
  }

  scale {
    type = "Standard"
  }
}

# write output blocks for each of the deployments
output "embedding" {
  value = azurerm_cognitive_deployment.embedding.name
}

output "gpt_35_turbo" {
  value = azurerm_cognitive_deployment.gpt_35_turbo.name
}

output "gpt_4" {
  value = azurerm_cognitive_deployment.gpt_4.name
}

output "gpt_4_32k" {
  value = azurerm_cognitive_deployment.gpt_4_32k.name
}