Hi, I Am Trying To Setup Multi-Node Training With Pytorch Distributeddataparallel. Ddp Requres A Launch Script With A Set Of Parameters To Be Run On Each Node. One Of These Parameters Is Master Node Address. I Am Currently Using The Following Scheme:

Unanswered

Thanks for the answer!

the idea is to use ClearML for the node allocation/scheduling and PyTorch DDP for the actual communication, is that correct ?

Yes, i basically plan to use ClearML as user-friendly cluster manager

Regrading Task pollution, when the master node is done, it can delete all child/cloned Tasks so it is easier on the eyes.

Nice idea, i will try it out!

  				
Posted 
	2 years ago

					More
				  		
  Report
		
					UnevenHorse85
				
					0
					 × 1

91 Views

0 Answers

2 years ago

one year ago