ghost
commented
11 years ago fixed now.
Closed rdevon closed 11 years ago
ghost
commented
11 years ago fixed now.
rdevon
commented
11 years ago Hey Nitish, sorry another question.
I'm slowly digging deeper into the code. Does l1decay need to be negative? It appears, on the surface, to be the case, but I have no idea what "edge.temp" is at that point (and I'm worried it's a catch-all variable), but I assume (the cuds code isn't clear on this detail), that if there is a target, the sign is relative to that. Why would I need to decay to anything but 0, and under what situations would I expect that to happen here?
thanks
-devon
On Apr 3, 2013, at 3:22 AM, Nitish Srivastava notifications@github.com wrote:
fixed now.
— Reply to this email directly or view it on GitHub.
Train Step: 0Traceback (most recent call last): File "../../../deepnet/deepnet/trainer.py", line 56, in
main()
File "../../../deepnet/deepnet/trainer.py", line 51, in main
model.Train()
File "/na/homes/dhjelm/CUDANET/deepnet/deepnet/neuralnet.py", line 646, in Train
losses = self.TrainOneBatch(step)
File "/na/homes/dhjelm/CUDANET/deepnet/deepnet/dbm.py", line 270, in TrainOneBatch
losses1 = self.PositivePhase(train=True, step=step)
File "/na/homes/dhjelm/CUDANET/deepnet/deepnet/dbm.py", line 131, in PositivePhase
self.ComputeUp(node, train=train)
File "/na/homes/dhjelm/CUDANET/deepnet/deepnet/dbm.py", line 108, in ComputeUp
if layer.pos_phase:
AttributeError: 'LinearLayer' object has no attribute 'pos_phase'