Jiaxin Shan

@feilengcui008 For normal case, it's true. The major blocker issue for us to adopt sts or deployment is we plan to enable autoscaler soon and we need to fine grain...

`compute template` is probably the best place to configure affinity and tolerations. Does it mean eventually we will have many kubernetes fields mapped to compute template? is there a better...

@gaocegege @kiukchung I can write a proposal on it. It needs some interface changes. We can have some discussion by end of month.

> FWIW torchelastic is being bundled with the PT docker (https://github.com/pytorch/pytorch/blob/master/Dockerfile#L55) in the next PT release That's a good news. operator will be only used for orchestration, it's user's responsibility...

@ajcann I think there's two separate issues. The first, I can confirm `--default-ulimit nofile=2048:8192` is still in accelerator AMI, I think I can either make a change to 65535 or...

@jinchihe @jlewi I think fairing, kale, or jupyter these are projects related to development experiences. It there a better abstraction of these projects and have a WG?

Thanks @PatrickXYS driving this. I think you should have write access to this repo. Non of these branches are active, please help remove them

@PatrickXYS Yeah, if some branches can not be deleted. I think it's fine for someone with permission to clean up later. Not a big deal

@zhenghuiwang This would be great. The next release you mean is 0.7? BTW, any examples for current release (0.6) we can try?

/cc @Tomcli @adrian555 @shawnzhu