If you need a classifier trained, but don't want to bother with the lengthy net configuration, these scripts configure a pre-trained version of the alexnet CNN performs well on many vision problems, and converges particularily quickly on academic figures. Note that caffe and dependencies must be installed prior scripts if ami not used.
The script is built to run on the following AMI:
TEMPLATEFORCNNTRAINING
...So if you don't use it, you are on your own for path configuration,etc.
To run:
-put dataset in PUTDATAHERE folder (This would be a directory of two(train-val) and a list of all the files with their class (integer) following their name) -edit allsetup.sh to point to the folder and select the number of classes -double check allsetup.sh -run ./allsetup.sh -if everything works, run ./runit.sh