2.5.0

2.5.0

• Add -AdditonalInstructions parameter to Start-ThreadRun.

2.4.0

2.4.0

• Add -LogProbs and -TopLogProbs parameter to Request-ChatCompletion.
• Add -Images and -ImageDetail parameter to Request-AzureChatCompletion.
  These parameters are currently acceptable only on the gpt-4-vision-preview model.

2.3.0

2.3.0

• Request-AzureImageGeneration now supports DALL-E 3 model.
• Change default api version of Azure OpenAI Service to 2023-12-01-preview
• Fix minor issues.

2.2.0

2.2.0

• Add Request-AudioSpeech function. It generates audio from the input text.

  PS C:\> Request-AudioSpeech -text 'Do something fun to play.' -OutFile 'C:\Output\text2speech.mp3' -Voice Alloy

• Add new functions for using the OpenAI Assistants API.

  To use Assistants with the new PSOpenAI cmdlets, please refer to this guide.
  Guide: How to use Assistants

  [!WARNING]
  The Assistants API is still in Beta. Specifications, usage, and parameters are subject to change without announcement.

  • Assistants: Get-Assistant, New-Assistant, Remove-Assistant, Set-Assistant
  • Threads: Get-Thread, New-Thread, Remove-Thread, Set-Thread
  • Messages: Get-ThreadMessage, Add-ThreadMessage
  • Runs: Get-ThreadRun, Start-ThreadRun, Stop-ThreadRun, Wait-ThreadRun, Receive-ThreadRun
  • Steps: Get-ThreadRunStep
  • Files: Get-OpenAIFile, Register-OpenAIFile, Remove-OpenAIFile, Get-OpenAIFileContent

• Change directory name from "Examples" to "Guides".

• Suppress verbose message about the organization-Id not found.

• Fix minor bugs.

2.1.0

2.1.0

Implements some updates that announced in OpenAI Dev Day 2023.
(New features such as Threads or Assistants does not implemented yet. We are working in progress.)

• Add / Rename / Remove some parameters to Request-ChatComplention.

  • Add : -Images, -ImageDetail, -Tools, -ToolChoice, -Format, -Seed
  • Remove : -Functions, -FunctionCall, -MaxFunctionCallCount
  • Rename : -InvokeFunctionOnCallMode to -InvokeTools

• Add new models to tab completions of Request-ChatComplention.

  • gpt-3.5-turbo-1106, gpt-4-1106-preview, gpt-4-vision-preview

• Add -Model parameter to Request-ImageGeneration. You can specifiy the model name. (dall-e-2 or dall-e-3)

• Add -Quality and -Style parameters to Request-ImageGeneration (for dall-e-3)

2.0.0

2.0.0

This is a major release that includes breaking changes.

• Add -ApiBase parameter to specify the base URL of the API endpoint.
  This is useful for using the OpenAI compatible API such like FastChat or LM Studio in a private environment.

  PS C:\> Request-ChatCompletion -Message 'Hello' -ApiBase 'https://localhost:8000/v1'

• Get/New/Remove-AzureOpenAIDeployments functions are removed.
• Remove -Token parameter from all functions. Use -ApiKey instead.
• OPENAI_TOKEN environment variable is deprecated. Use OPENAI_API_KEY instead.
• The default model for Request-TextCompletion is changed to gpt-3.5-turbo-instruct from text-davinci-003.
• Add -Format parameter to Request-Embeddings function. This parameter can be used to specify the format of the returned embeddings.

1.15.2

1.15.2

• Fix various issues about Function calling.

1.15.1

1.15.1

• Fix various issues about Function calling.

1.15.0

1.15.0

• Add new whisper functions for Azure OpenAI Service
  • Request-AzureAudioTranscription
  • Request-AzureAudioTranslation
• Remove the upper limit on the -MaxTokens parameter. (The maximum number of tokens depends on the model used)
• Add support for Function calling method on Azure OpenAI Service.
• Change default api version of Azure OpenAI Service to 2023-09-01-preview

1.14.3

1.14.3

• Add a new gpt-3.5-turbo-instruct model to tab completions of Request-TextCompletion.