keras nn script was working fine and all of a sudden i started getting this error. http://pastebin.com/7g0ubynq theano gpu test works @fchollet
-
@fchollet yes. it was working fine, but when i increased he number of features, i started getting this error. it works fine for less data.
@abhi1thakur Sounds like a Theano bug, raise it with the Theano devs. Also you can try using TensorFlow instead
-
-
@fchollet unfortunately tensorflow needs cuda compatibility of 3.5+ . cannot run it on amazon ec2's grid gpu :( -
@abhi1thakur However convnets in TF are fairly fast on CPU (goes to full utilization). It wouldn't be crazy to run TF on large C instances.
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.