Version 0.3.1

• Adding GPT-4 models (with 8k and 32k context)
• Adding an optional display (header) file name to OpenAIService.uploadFile function. If not specified a full path is used as a file name instead.
• Removing dependence on play.shaded...HttpHeaderNames in MultipartWritable
• Removing the obsolete OpenAIService's function - retrieveModelVersions

Version 0.3.0

• New OpenAIService functions/endpoints: createChatCompletion (ChatGPT API), and audio stuff - createAudioTranscription and createAudioTranslation
• New OpenAIServiceStreamedExtra function/endpoint: createChatCompletionStreamed
• New models: whisper-1, gpt-3.5-turbo, and gpt-3.5-turbo-0301
• DeleteResponse(s) changed from Enumeration to a sealed trait with case classes.
• Command and Tag enumerations relocated to a separated file + PEP#Value (Command) and PT#Value (Tag/Param) in WSRequestHelper and OpenAIServiceImpl functions simplified/generalized to PEP and PT.

Version 0.2.0

• Streaming support for createCompletion and listFineTuneEvents
• Introducing OpenAIServiceFactoryHelper to be used for new OpenAIServiceStreamedFactory
• Adding types to JsonFormats (required by Scala-3)
• Making a few attributes optional (e.g., FineTuneJob->events)

Version 0.1.1

• The attribute completion_tokens of UsageInfo (in TextCompletionResponse) made optional... most likely due to an API change
• A few edits in README (e.g. ref to a Medium article)

Version 0.1.0

• Sonatype publish process fixed
• Closing WS connection with a system shutdown hook