parser.add_argument("checkpoint_file", help='Path to single model checkpoint (.pkl) file.') args = parser.parse_args() checkpoint_file = args.checkpoint_file fold = int(args.split) dataset_path = os.path.join(data_paths.tfd_data_path, 'npy_files/TFD_96/split_'+str(fold)) print 'Checkpoint: %s' % checkpoint_file print 'Testing on split %d\n' % fold # Load model model = SupervisedModel('evaluation', './') # Load dataset supervised_data_loader = SupervisedDataLoader(dataset_path) test_data_container = supervised_data_loader.load(2) test_data_container.X = numpy.float32(test_data_container.X) test_data_container.X /= 255.0 test_data_container.X *= 2.0 # Construct evaluator preprocessor = [util.Normer3(filter_size=5, num_channels=1)] evaluator = util.Evaluator(model, test_data_container, checkpoint_file, preprocessor) # For the inputted checkpoint, compute the overall test accuracy accuracies = [] print 'Checkpoint: %s' % os.path.split(checkpoint_file)[1] evaluator.set_checkpoint(checkpoint_file)
parser.add_argument("checkpoint_dir", help='Folder containing all .pkl checkpoint files.') args = parser.parse_args() checkpoint_dir = args.checkpoint_dir fold = int(args.split) dataset_path = os.path.join(data_paths.tfd_data_path, 'npy_files/TFD_96/split_'+str(fold)) print 'Checkpoint directory: %s' % checkpoint_dir print 'Testing on split %d\n' % fold # Load model model = SupervisedModel('evaluation', './') # Load data supervised_data_loader = SupervisedDataLoader(dataset_path) val_data_container = supervised_data_loader.load(1) val_data_container.X = numpy.float32(val_data_container.X) val_data_container.X /= 255.0 val_data_container.X *= 2.0 # Construct evaluator preprocessor = [util.Normer3(filter_size=5, num_channels=1)] checkpoint_file_list = sorted( glob.glob(os.path.join(checkpoint_dir, '*.pkl'))) evaluator = util.Evaluator(model, val_data_container, checkpoint_file_list[0], preprocessor) # For each checkpoint, compute the overall val accuracy accuracies = []
f.write(str(pid)+'\n') f.close() # Load model model = SupervisedModel('experiment', './', learning_rate=1e-2) monitor = util.Monitor(model, checkpoint_directory='checkpoints_'+str(train_split), save_steps=1000) # Add dropout flag to fully-connected layer model.fc4.dropout = 0.5 model._compile() # Loading TFD dataset print('Loading Data') supervised_data_loader = SupervisedDataLoader( os.path.join(data_paths.tfd_data_path, 'npy_files/TFD_96/split_'+str(train_split))) train_data_container = supervised_data_loader.load(0) val_data_container = supervised_data_loader.load(1) test_data_container = supervised_data_loader.load(2) X_train = train_data_container.X y_train = train_data_container.y X_val = val_data_container.X y_val = val_data_container.y X_test = test_data_container.X y_test = test_data_container.y X_train = numpy.float32(X_train) X_train /= 255.0 X_train *= 2.0
if args.which_set == 'train': set_num = 0 elif args.which_set == 'val': set_num = 1 else: set_num = 2 print 'Checkpoint: %s' % checkpoint_file print 'Evaluating on split %d' % fold print 'Using %s set\n' % args.which_set # Load model model = SupervisedModel('evaluation', './') # Load dataset supervised_data_loader = SupervisedDataLoader(dataset_path) data_container = supervised_data_loader.load(set_num) data_container.X = numpy.float32(data_container.X) data_container.X /= 255.0 data_container.X *= 2.0 print data_container.X.shape # Construct evaluator preprocessor = [util.Normer3(filter_size=5, num_channels=1)] evaluator = util.Evaluator(model, data_container, checkpoint_file, preprocessor) # For the inputted checkpoint, compute the overall accuracy accuracies = [] print 'Checkpoint: %s' % os.path.split(checkpoint_file)[1]