search

hidasib / GRU4Rec

GRU4Rec is the original Theano implementation of the algorithm in "Session-based Recommendations with Recurrent Neural Networks" paper, published at ICLR 2016 and its follow-up "Recurrent Neural Networks with Top-k Gains for Session-based Recommendations". The code is optimized for execution on the GPU.
Other
754 stars 223 forks source link

How long does it take to run this program? #17

Closed zoltar9264 closed 6 years ago

zoltar9264 commented 6 years ago

Hello, I am a beginner of python and Neural Network 。 I run this code in a 32 cores cpu without GPU. "run_rsc15.py" has run amost 24 hours. So I want know how long it will take to run "python run_rsc15.py". by the way, memory of the server is 128G. think you very much !

zoltar9264 commented 6 years ago

finally,this program finished with 51 hours.

hidasib commented 6 years ago

Please note that training is optimized or GPU and not CPU. On GPU training should finish under 6 hours (depending on the parameters you use it can be much faster). If you plan to stik to CPU I'd suggest using the cross-entropy loss as it uses less of the GPU specific speed-ups that make the network train slower on CPU. (Note that opimal hyperparameters for the cross-entropy loss are different than for BPR-max.)

Additional speed-up for GPU will be published shortly.

CrisMon-01 commented 4 years ago

@hidasib How can I see if it's running on GPU? I have run the gpu_ops.py rogram. After that i hae run the run_rsc15.py but i see on my machine (ubuntu with a nvidia 960M) with the command: nvidia-smi -l there is no python programm runnign n that