Memory exhausted error about pygta5

[frossaren]

Solved by adding batch_size in the model.fit row

[johndpope]

see 20 mins in for a config you can set to avoid tensorflow hogging ram. https://blog.altoros.com/video-and-slides-from-the-tensorflow-london-meetup-march-31-2016.html

@Sentdex - I think this is related to your memory hogging problem on your 7/8/9 video.

[johndpope]

I think you want to call tf.reset_default_graph() somewhere. although it's not clear where.
you will likely need to
import tensorflow as tf
the code should expose this.

[frossaren]

@johndpope Where am i supposed to use this config? Tried but keeps getting the same error

[johndpope]

sorry @frossaren - it's a stab in the dark.
watching the tutorials - I understood that the machine learning was using tensorflow.
but searching through code - it doesn't appear - although there is a reference to tflearn.
if you can get the tensorflow session - then you should be able to configure the gpu memory options.

[frossaren]

Well then i am completely stuck. Will keep on googleing and see if i can find something useful and update here if i do.

[johndpope]

you should do a PR. when you help another you automatically help yourself. it's like multi tasking.

[frossaren]

I will, but i will see if it affected the performance first and If it gets better with a higher batch size. Thanks for the reminder Den 1 aug. 2017 03:14 skrev John Pope <[email protected]>: you should do a PR. when you help another you automatically help yourself. it's like multi tasking. — You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub<#59 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/ATxZZxSuS-Z3jaLf7k4HHOGPKg3TIXaUks5sTnwRgaJpZM4OcfKE>.

[frossaren]

performance seemes to be unaffected so i will do a pr