Dense V-Net not able to process input image

[Lazy-Propagator]

ValueError: input tensor's spatial dimensionality not not compatible, please tune the input window sizes: lambda x: x % modulo == 0)

I receive this error when I try to run Dense V-Net on a mhd file, when I am specifying the dimensions in spatial_window_size, which are (512, 512, 194).

When I don't specify anything, I get an error that: ValueError: Unknown output window size for input image image Can someone please point out what is wrong? I've attached the config file for reference Thank you ############################ input configuration sections [ct] path_to_search = ./data/dense_vnet_abdominal_ct/ filename_contains = CT spatial_window_size = (512, 512, 194)

interp_order = 1

axcodes=(A, R, S)

[label] path_to_search = ./data/dense_vnet_abdominal_ct/ filename_contains = Label spatial_window_size = (512, 512, 194)

interp_order = 0

axcodes=(A, R, S)

############################## system configuration sections [SYSTEM] cuda_devices = "" num_threads = 1 num_gpus = 1 model_dir = models/dense_vnet_abdominal_ct queue_length = 36

[NETWORK] name = dense_vnet

batch size 1 for inference

batch size 6 for training

batch_size = 1

volume level preprocessing

volume_padding_size = 0 window_sampling = resize

[TRAINING] sample_per_volume = 1 lr = 0.001 loss_type = dense_vnet_abdominal_ct.dice_hinge.dice starting_iter = 0 save_every_n = 1000 max_iter = 3001

[INFERENCE] border = (0, 0, 0) inference_iter = 3000 output_interp_order = 0 spatial_window_size = (512, 512, 194) save_seg_dir = ./segmentation_output/

############################ custom configuration sections [SEGMENTATION] image = ct label = label label_normalisation = False output_prob = False num_classes = 9

[wyli]

Hi @Lazy-Propagator the window size should be divisible by 8