Skip to content

caffepredict.sh

Chris Churas edited this page Oct 15, 2018 · 6 revisions

This is an internal script run by [runprediction.sh][runprediction.sh] that runs caffe prediction on a specific model within the directory created by runprediction.sh

Usage:

usage: caffepredict.sh [-h]
                      model augimagesdir outputdir

              Version: 1.6.0

              Runs caffe on CDeep3M model specified by model argument 
              to perform training. The trained model will be stored in
              <model>/trainedmodel directory
              Output from caffe will be redirected to <model>/log/out.log
              When script completes <outputdir>/DONE file will be created
              with last line containing 0 upon success. 
    
positional arguments:
  model                Path to .caffemodel file or directory with caffe
                       model. If later then the latest is used.
  augimagesdir         Directory path with prefix containing the 16 .h5
                       files that end with v#.h5. This data would have been
                       created PreprocessImageData.m
  outdir               Destination directory to write output. Will be
                       created if it does not exist.
optional arguments:
  -h, --help           show this help message and exit
  --gpu                Which GPU to use, can be a number ie 0 or 1 or
                       all to use all GPUs (default all)

Example usage:

caffepredict.sh ~/sbem/mitochrondria/xy5.9nm40nmz/30000iterations_train_out ~/augimages ~/predictout
Clone this wiki locally