Devis Lucato

thanks @Stevenic - we're looking into this area while working on increasing the pipeline flexibility. The change should be straightforward, but we just need to make sure it doesn't conflict...

I think the work on typescript has taken a break and we haven't heard about similar scenarios elsewhere, so for now I'm closing this. We'll pick up the work once...

thanks @luisquintanilla, yes, we are familiar with the package and plan to use the official library eventually. There are a few aspects to consider, such as batching, HTTP retry logic,...

@luisquintanilla work done, v0.12 now uses the Azure SDK for text, chat and embeddings. We still have a custom client for DALL-E2. Please let us know if there are updates,...

@RogerBarreto some conflicts to address. There's also another PR touching the HTTP layer, so could happen again, in one or the other, depending on the order these will be merged

100% agree with the premise and in fact I recently started drafting a doc to think through this holistically. I was trying to integrate a separate model and noticed the...

Using GPT turbo is reasonably simple using a connector. I think most of the friction is about persisting the chat history object inside the context, with a continuous serialization/deserialization, which...

quick update: work is in progress, here's the pull request adding ChatGPT and DallE: https://github.com/microsoft/semantic-kernel/pull/161

@andreracz thanks for the report. Could you please provide some examples about the prompts where you noticed the difference? Also, which model did you use for these tests? Usually mixing...

just adding an example, using text-davinci-003 and mixing languages: