Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

SkyRL brings Tinker to your GPUs (2025)

novasky-ai.notion.site

Post reply on HN
Print view View original

1–7 of 7 posts

Re: SkyRL brings Tinker to your GPUs (2025)

#2

Post by pcmoritz »

We are excited about the Tinker API, it exposes the primitives forward_backward, optim_step, sample and checkpoints of an LLM as a REST API, which can be used to implement pretty much arbitrary training recipes while hiding all the infra challenges of running the model and also abstract the underlying accelerator with very small surface area. We hope it can emerge as a standard. If you have any feedback for our open-source implementation we would love to hear about it!

Re: SkyRL brings Tinker to your GPUs (2025)

#4

Post by veryluckyxyz »

pcmoritz wrote: post #2

We are excited about the Tinker API, it exposes the primitives forward_backward, optim_step, sample and checkpoints of an LLM as a REST API, which can be used to implement pretty much arbitrary training recipes while hiding all the infra challenges of running the model and also abstract the underlying accelerator with very small surface area. We hope it can emerge as a standard. If you have any feedback for our open-…

Consider fixing your “published date” on the site.

Re: SkyRL brings Tinker to your GPUs (2025)

#6

Post by pcmoritz »

veryluckyxyz wrote: post #4
pcmoritz wrote: post #2

We are excited about the Tinker API, it exposes the primitives forward_backward, optim_step, sample and checkpoints of an LLM as a REST API, which can be used to implement pretty much arbitrary training recipes while hiding all the infra challenges of running the model and also abstract the underlying accelerator with very small surface area. We hope it can emerge as a standard. If you have any feedback for our open-…

Consider fixing your “published date” on the site.

Thanks, fixed!
Post reply on HN