Skip to content

Releases: jodendaal/OpenAI.Net

v1.0.21

10 Nov 14:05
3598c4f
Compare
Choose a tag to compare
  • Dall-E 3 support for ImageGenerationRequest, added properties for :
    • Model
      • ImageModelOptions
    • Quality
      • ImageQualityOptions
    • Style
      • ImageStyleOptions
  • Json Mode for ChatCompletionRequest, added property for
    • ResponseFormat
      • ChatResponseFormat

v1.0.20

25 Mar 10:00
5a1ee34
Compare
Choose a tag to compare
  • TextEdit Service Removed
    • Breaking Change. See below on how to switch over if you where using this.

The v1/edits API has been removed by OpenAI and no longer works.

The following models have also been deprecated and no longer work. These have also been removed.

text-davinci-edit-001
audio-transcribe-001
code-cushman-001
code-davinci-002
code-davinci-edit-001
text-davinci-edit-001
text-davinci-insert-001
text-davinci-insert-002

OpenAI suggests using the ChatCompletion endpoint with a query like this.

[Test]
public async Task TetEditReplacement()
{
    var messages = new List<Message>
    {
        Message.Create(ChatRoleType.System, "You are a spell checker. Fix the spelling mistakes"),
        Message.Create(ChatRoleType.User, "What day of the wek is it?"),
    };

    var response = await OpenAIService.Chat.Get(messages, o => {
        o.MaxTokens = 1000;
    });

    Assert.That(response.IsSuccess, Is.True);
    Assert.That(response.Result.Choices[0].Message.Content, Is.EqualTo("What day of the week is it?"));
}

v1.0.19

20 Mar 22:25
c0b2ad3
Compare
Choose a tag to compare
  • Added PresencePenalty property to TextCompletionRequest.

v1.0.18

18 Mar 11:32
05fd472
Compare
Choose a tag to compare
  • Updated LogitBias Type from object to Dictionary<string,int>
  • Updated and added some additional Integration tests
  • Updated Readme to include Chat and Audio examples.

v1.0.17

14 Mar 20:40
13ad2dc
Compare
Choose a tag to compare
  • Added Support for Chat API.
  • Added Support for Audio API.

v1.0.15

22 Jan 18:05
Compare
Choose a tag to compare

v1.0.15

Corrected bug with FineTune Create response erroring if LearningRateMultiplier or BatchSize was null.

v1.0.14

08 Jan 20:17
Compare
Choose a tag to compare

Updated to allow implicit return and initialisation of OperationResult.
This allows returning the response without needing to specify OperationResult<,> as the return type or needing to do result.Result.

e.g

private async Task<ModelInfo> GetResponse()
{
    var result = await OpenAIService.Models.Get(ModelTypes.TextDavinciEdit001);
    if (result.IsSuccess)
    {
        return result;
    }
    else
    {
        throw new Exception(result.ErrorMessage);
    }
}

rather than

private async Task<ModelInfo> GetResponse()
  {
      var result = await OpenAIService.Models.Get(ModelTypes.TextDavinciEdit001);
      if (result.IsSuccess)
      {
          return result.Result;
      }
      else
      {
          throw new Exception(result.ErrorMessage);
      }
  }

rather than

 private async Task<ModelInfo> GetResponse()
  {
      var result = await OpenAIService.Models.Get(ModelTypes.TextDavinciEdit001);
      return result.Result;
  }
``

v1.0.13

29 Dec 23:04
2276957
Compare
Choose a tag to compare

v1.0.13

  • Updated to include XML comments file in nuget package

v1.0.12

27 Dec 20:53
310a849
Compare
Choose a tag to compare

v1.0.12

  • No code changes in this release. Just a release to update nuget package tags.

v1.0.11

27 Dec 20:36
889a63a
Compare
Choose a tag to compare

v1.0.11

  • Added extensions methods for ImageInfo to assist in converting from base64 to FileContentInfo.
  • Added extensions methods for strings to assist in converting from base64 to byte[] and FileContentInfo.
  • Corrected spelling mistake in ImageExtensions Generate method.
  • Added logo on Readme.
  • Added logo for nuget package.