search

AI4HealthUOL / SSSD

Repository for the paper: 'Diffusion-based Time Series Imputation and Forecasting with Structured State Space Models'
MIT License
270 stars 47 forks source link

How to train in parallel？ #7

Closed YangYu-NUAA closed 1 year ago

YangYu-NUAA commented 1 year ago

Hello, sorry to bother you again, but I would like to explore how to make this model run on multiple GPUs. When I try to set the net (S4) in parallel (using net = nn.DataParallel(net, device_ids=[0,1])), it always reports an error. Thank you!

juanlopezcode commented 1 year ago

Hello @YangYu-NUAA, unfortunatelly, I never trained the model in parallel, however, as an act of good will I could provdide you some hint to debug it (if I get one), but I would need that you paste here the error traceback

YangYu-NUAA commented 1 year ago

Hello @YangYu-NUAA, unfortunatelly, I never trained the model in parallel, however, as an act of good will I could provdide you some hint to debug it (if I get one), but I would need that you paste here the error traceback

Sorry but it is weird，i cannot recurrence that bug when i usenet = nn.DataParallel(net, device_ids=[0,1]). lol now it can run in parallel.