Michael Doyle

Hey @nechmads can you share an example prompt and/or code. I think I see what you are likely running into - but want to verify. Vertex AI also supports Anthropic,...

It's true :) I wrote that in the description. From a UX perspective, we might still want to do it. BUT, if not. Let's close this out. Let me know.

Apparently multiple images is supported natively in the genai sdk. This is working for Go now.

Agreed - I suspect there is a brief discussion that needs to happen on "how" we want this to be supported in the core library, and then implementation details fall...

This is an old issue, but in case anyone runs into it, I ended up fixing this problem for myself (among many other improvements). Since this project is not being...

**hello.history.prompt** ```yaml --- model: googleai/gemini-1.5-flash config: maxOutputTokens: 2048 temperature: 0.6 topK: 16 topP: 0.95 safetySettings: - category: HARM_CATEGORY_HATE_SPEECH threshold: BLOCK_ONLY_HIGH - category: HARM_CATEGORY_DANGEROUS_CONTENT threshold: BLOCK_ONLY_HIGH - category: HARM_CATEGORY_HARASSMENT threshold: BLOCK_ONLY_HIGH...

I think that's the right approach. Been swept up in some other things, but will try and knock this out today.

@pavelgj I got this going for the `generate` case. I wasn't exactly sure what to do for `prompt.stream()`, though, since `generateStreaming` is sync while `runInNewSpan` returns a Promise.

@schnecle same applies to the generate util as well, correct? maybe we should do a separate pass on metrics. I think theres a couple things there.

Hey @ssbushi were there other fixes needed besides https://github.com/FirebasePrivate/genkit-ui/pull/1315? Or can we close this one out?