gmKeshari

@anupdsouza Thank you reporting this issue. This looks like an intermittent error and should work now. This repository is for issues related to Swift SDK client bugs or improvements and...

Hi @kashish63, Are you using Json mode in your code?? Can you provide your code, so i can reproduce the issue from my end if possible?

Hey @kashish63, You can try below suggestions in your code : Use proper error handling, include fallback responses, check external API reliability, improve your prompt. **Robust Error Handling:** Implement comprehensive...

This problem is common with LLMs, they can sometimes hallucinate or deviate from instructions, even with low temperature and top_p settings. Here I can suggest few changes : 1. More...

Hi @imperorrp, Thanks for flagging. For the first issue "Tuning example with new sdk", team is working to shift everything to new sdk, i will follow up on this. For...

Tuning might be available through Vertex AI only going forward. Tuning has been removed from Google AI Studio, that could be the reason it's not updated to the new sdk.

Hi @jamesblackwell, Recitation as finish_reason you get when model tries to generate copyright content. You can find more info [here](https://ai.google.dev/gemini-api/docs/troubleshooting?lang=python#recitation-issue). To fix this, try to make prompt / context as...

Hey @jamesblackwell, Thanks for providing the more context. I will raise this with our internal team for workaround.

Hi @seungduk-yanolja. I found a similar issue on the Developers forum : https://discuss.ai.google.dev/t/json-mode-internal-server-error-500/38123 I will escalate this. Thanks!

Hi @lennijusten , Have you been facing this issue since the beginning ? I mean is it like you were getting the response earlier and now facing this issue? Try...