neural_graph_collaborative_filtering icon indicating copy to clipboard operation
neural_graph_collaborative_filtering copied to clipboard

Published 20 hours ago • verified publisher icon xiangwang1223

how to test the performence in the test dataset

Open namenotexist opened this issue 5 years ago • 8 comments

namenotexist avatar Aug 10 '19 04:08 namenotexist

why the training process so slow

namenotexist avatar Aug 11 '19 05:08 namenotexist

In my experience, u need a GPU. Tested on 2080ti, 20s for 1 epoch while without GPU, it took 800s 1 epoch. And by the way, 99% the training time is used in sample process.

wanjunhong0 avatar Aug 12 '19 06:08 wanjunhong0

why did i often occur "loss is nan"

namenotexist avatar Aug 13 '19 14:08 namenotexist

You can try the following code to modify the BPR loss, in order to avoid NAN:

loss = tf.reduce_sum(tf.nn.softplus(-(pos_result - neg_result)))

xiangwang1223 avatar Aug 14 '19 01:08 xiangwang1223

you mean tf.reduce_mean?

namenotexist avatar Aug 14 '19 12:08 namenotexist

why did i can only use 149M GPU memory, which has a full memory of 10GB

namenotexist avatar Aug 14 '19 14:08 namenotexist

I have no idea... Have you installed the tensorflow-gpu successfully?

xiangwang1223 avatar Aug 15 '19 03:08 xiangwang1223

I have no idea... Have you installed the tensorflow-gpu successfully? i fix it . the reason is that i install tensorflow-gpu 1.14.0,when i change it to tensorflow-gpu 1.8.0, it works

namenotexist avatar Aug 15 '19 03:08 namenotexist