CaffeConTroll
CaffeConTroll copied to clipboard
HazyResearch
Will the techniques for speeding up neural networks used in CaffeConTroll (lowering techniques and batching) work well for mobile devices?
How to calculate flops spent for forward/backward pass? For example we have two machines one have some old GPU and another one - modern GPU, so when we train some...
fail to run googlenet, got following error message: Constructing CONV bridge Sharing data pointers with previous bridge (no input cube allocation or data copy) Gradient updates will happen on the...
Removed "azureuser" char to use $$USER instead to remove the common error "cblas.h: No such file or directory"
The tests point to hard-coded data path like `/home/ubuntu/data/ilsvrc12_train_1000_lmdb`, which seems to be available only on those VM images. Would it be easy to host the data files somewhere and...
Looks like the commands in README and the code itself are out of sync.
sgeadmin@master:~/CaffeConTroll$ ./caffe-ct train caffe_inputs/lenet_solver.prototxt Start writing 60000 to toprocess.bin... Finished writing images to toprocess.bin... Constructing CONV layer with Grouping = 1 (# Input Grouping=1) Constructing MAXPOOLING (# Input Grouping=1) Constructing...
