Tried using the new gpt-3.5-turbo model. Now the censoring is built-in into the model itself. ๐คฆ
If you're using da-vinci model, don't migrate. The saving isn't worth it.
@antirez@liuliu@drawthingsapp Can you add int8 convrot implementation please, it is way more memory efficient and M5+ series will also benefits from it.
@OfficialLoganK How do you define reliability? Is there an actual SLA for Gemini API for โflexโ and โpriorityโ?
Gemini API has been quite disappointed with lot of 5xx resource overload lately. I want it to success but I canโt recomment serious projects to use it right now.
@OfficialLoganK Hi Logan, can you pls add markdown file (.md) as support file type in AI Studio. Currently when selecting a markdown file, it say: file type unsupported.
@tdinh_me I doubt the โmaking bankโ part ๐ฌ. Theyโre paying Anthropic, OpenAI for the LLM calls.
And if people make a lot of requests + big codebase, it might even lost them money.
@alexalbert__ Hi @alexalbert__ , TailwindCSS doesnโt work in Artifact if itโs added via CDN. Can you raise this to the team please? Itโs a very popular CSS library.
Hi @AnthropicAI , TailwindCSS doesnโt work for Artifact. Maybe due to some iframe misconfiguration that prevent loading CSS from a CDN.
Can you please have a look?
To reproduce, just ask Sonnet3.5 to generate a simple page with TailwindCSS.
@BHolmesDev Astro is nice, has been using it for my pet projects. Is there a way to setup cron job for it, with NodeJS adapter? I canโt find any guide
@williamfalcon It isnโt that easy to use though. For Colab, I already has my Gmail account and it works.
Try Lightning, they require registering, andโฆ what, I need to wait 2-3 days for my email to be verified? Can we just send a confirmation email instead ??
@emollick Can you try asking it to build a full-stack app where it need a DB to store data? Most of the examples seems to be for client-side apps only.