Re: [DISCUSS] Horn Incubation Proposal

2015-08-30 Thread Edward J. Yoon
Hi, If there are no more comments/concerns, I plan to call for a vote soon. Thanks! On Wed, Aug 26, 2015 at 9:37 PM, Edward J. Yoon wrote: > More diverse volunteers for Horn project will be added, from > cldi-kaist team who is ranked 7th in the ImageNet ILSVRC last year. > We're also seeking one

Re: [DISCUSS] Horn Incubation Proposal

2015-08-26 Thread Edward J. Yoon
More diverse volunteers for Horn project will be added, from cldi-kaist team who is ranked 7th in the ImageNet ILSVRC last year. We're also seeking one more mentor (instead of me). Thanks. P.S., I'm CC'ing current volunteers. If you're still not on the list please subscribe so you can see what's

Re: [DISCUSS] Horn Incubation Proposal

2015-08-20 Thread Edward J. Yoon
> multiple worker groups for asynchronous training---data parallelism; and > multiple workers in one group for synchronous training---model parallelism. So, it's basically execution of the multiple asynchronous BSP (Bulk Synchronous Parallel) jobs. This can be simply handled within only single BSP

Re: [DISCUSS] Horn Incubation Proposal

2015-08-20 Thread ooibc
Hi, I am an initial committer of Apache(incubating) SINGA (http://singa.incubator.apache.org/) Both SINGA and the proposal follow the general parameter-server architecture: workers for computing gradients; servers for parameter updating. SINGA has implemented the model and data parallelism