Self-Driving-Car-3D-Simulator-With-CNN OOM

OOM

Open PandaWhoCodes opened this issue 7 years ago • 6 comments

Resource exhausted: OOM when allocating tensor with shape[]

Nvidia 2GB 1050Ti What is your setup?

Sep 30 '17 20:09 PandaWhoCodes

ResourceExhaustedError (see above for traceback): OOM when allocating tensor with shape[512] [[Node: dense_1/bias/Assign = Assign[T=DT_FLOAT, _class=["loc:@dense_1/bias"], use_locking=true, validate_shape=true, _device="/job:localhost/replica:0/task:0/gpu:0"](dense_1/bias, dense_1/Const)]]

Sep 30 '17 20:09 PandaWhoCodes

Sorry, I did not see this until today. Its probably because you're not getting the right screenshot area? Are you sure you have the right section of the screen?

Oct 11 '17 23:10 sagar448

Yes, Im sure

Oct 16 '17 06:10 PandaWhoCodes

I'm having the same issue here. First with the batch size of 32 I was getting cudnn_status_alloc_failed on model.predict() The I resized the batch size to 8 and the first error has gone but a this new one has showed up

Not even on a lower batch size of 4 it runs

My setup is the GTX 1060 3GB, Intel Core i5 7400 and 12GB RAM

Jul 30 '18 01:07 devdouglasferreira

Has the issue been resolved or it still persists?

Sep 02 '18 04:09 sagar448

Yes it still does

Sep 19 '18 04:09 PandaWhoCodes

Self-Driving-Car-3D-Simulator-With-CNN Self-Driving-Car-3D-Simulator-With-CNN copied to clipboard

OOM

Self-Driving-Car-3D-Simulator-With-CNN
Self-Driving-Car-3D-Simulator-With-CNN copied to clipboard