I worked with recurrent nets for dynamic system prediction for an undergrad thes... | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

		alkonaut on Feb 4, 2019 \| parent \| context \| favorite \| on: The Unreasonable Effectiveness of Recurrent Neural... I worked with recurrent nets for dynamic system prediction for an undergrad thesis in ‘02 and we really struggled to train them on the CPUs we had. They had probably fewer than 100 parameters.

minimaxir on Feb 4, 2019 [–]

In the GPU space, the new CuDNN RNNs are almost as fast as running MLPs on GPUs.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact