Maksim Khadkevich
Maksim Khadkevich
@yuzisun , you got me right. We are running multi-node inference on large NLP models.
@yuzisun we are running multi node inference following this tutorial https://github.com/triton-inference-server/fastertransformer_backend/blob/v1.1/docs/gptj_guide.md using both plain `mpirun` and `srun` wrapper. Now the goal is be able to find the closest alternative that...
pull requests are welcome
@hubin858130 , can you create a pull request?
is there a timeline for 2.0?
@parano , what is the recommended way to deploy models to yatai given that bento 1.2 is not supported yet. Should we stick to bento 1.1?
@milend , is there a roadmap with milestones for closing the gaps between internal and open-sourced versions of buck2? Having a matrix containing a list of platforms that are currently...
@nikg4 , is it a linear ticket number where the progress is tracked? Were there any updates on the k8s support plans?
is there any enthusiasm from the open-source community to keep driving this effort? were there any forks where the community keeps building features?