Hi I have tried to run AMPlify successfully on CPU only, but slow. Now trying to run AMPlify with tensorflow-gpu (we have 4 x Nvidia A800), it ran into this error consistently in prediction after model loaded.
No matter how I tried to set many parameter in session config, e.g.
gpu_options.allow_growth =True, (in fact, if false, one GPU mem is not peaked, only one GPU card is used), see below images
PS. my input peptide has two aa sequences only
Being rather new to tensorflow, any chance some one may help advice what 's gone wrong
Hi I have tried to run AMPlify successfully on CPU only, but slow. Now trying to run AMPlify with tensorflow-gpu (we have 4 x Nvidia A800), it ran into this error consistently in prediction after model loaded.
No matter how I tried to set many parameter in session config, e.g. gpu_options.allow_growth =True, (in fact, if false, one GPU mem is not peaked, only one GPU card is used), see below images
PS. my input peptide has two aa sequences only
Being rather new to tensorflow, any chance some one may help advice what 's gone wrong