Skip to content

Conversation

@MarkHoath
Copy link

I created another init() function which takes the old APIKEY and passes it to the new Config structure.

Because Images are handled by different libraries in IOS and MacOS, there is no error handling.

Images have to be the correct size (256x256, 512x512, 1024x1024 and less than 4MB)

Images must be png files and passed as image.pngData(). (same for the Mask in Image Edit endpoint)

For Editing, if no mask is supplied (its optional) then the image/png file MUST have an alpha channel. (RGBA).

Any of the above Failures result in "Success" results, but "Errors" in the response, which aren't handled by the Decoder.

MarkHoath and others added 30 commits July 16, 2023 16:42
Fixes adamrushy#100 `sendChat is failing because ChatMessage has extra property`
Additional properties are not allowed ('id' was unexpected)
The current model enum and methods that have models as parameters could easily cause errors as any model could be chosen for any of those methods. By adding the new `OpenAIEndpointModel` type that corresponds to the compatibility list in the OpenAI docs, it’s much easier to select the fitting model. The new type does contain duplicate model strings, which is intended to make it easier to maintain. Backward compatibility is assured by marking the previous methods as deprecated. https://platform.openai.com/docs/models/model-endpoint-compatibility
@MarkHoath
Copy link
Author

This update includes the Jan 2024 Endpoints and Models including gpt4-turbo.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

7 participants