HackerTrans
TopNewTrendsCommentsPastAskShowJobs

essekar

no profile record

Submissions

[untitled]

1 points·by essekar·mese scorso·0 comments

[untitled]

1 points·by essekar·2 mesi fa·0 comments

[untitled]

1 points·by essekar·2 mesi fa·0 comments

[untitled]

1 points·by essekar·2 mesi fa·0 comments

[untitled]

1 points·by essekar·2 mesi fa·0 comments

Ramen at 8. Run after work. Movie tonight. Drop the plan. See who's innn

innn.life
7 points·by essekar·2 mesi fa·2 comments

NVSX: Adopt NVSentinel the Easy Way

github.com
1 points·by essekar·2 mesi fa·0 comments

Show HN: nvsx – A hook layer for NVSentinel GPU fault remediation

github.com
1 points·by essekar·3 mesi fa·1 comments

Singularity Is Around?

1 points·by essekar·4 mesi fa·0 comments

comments

essekar
·2 mesi fa·discuss
right now, I am scoping it to nearby people; within X miles distance. Saying strangers look fishy - might have to add more verifications.

How open would you be to trying go out?
essekar
·2 mesi fa·discuss
anyone has any videos?
essekar
·3 mesi fa·discuss
lol, that is what we all needed. AI Boss
essekar
·3 mesi fa·discuss
anything tbh. as long as you have runbook - you can try to automate actions through nvsx; it sits on top of NVSentinel. restarting will work mostly for smaller jobs - distributed training, pretty common will need more fault tolerant methods to continue rather than just restarting.
essekar
·3 mesi fa·discuss
[dead]
essekar
·4 mesi fa·discuss
One of the feedbacks we got while testing was - we might even reduce the duration of checkpointing. Which was a huge insight.

We are testing out & would love more collaboration from the community - if your team is running training jobs, hit us up.