Questions about test about asyndgan HOT 3 CLOSED

Please provide the error msg and the command?

from asyndgan.

python save_syn.py --dataroot /home/OYSK/codes/AsynDGAN/test_demo --name AtoB --model asyndgan --netG resnet_9blocks --direction AtoB --dataset_mode brats --results_dir results/test_asyndgan
----------------- End -------------------
build key:0
dataset [BratsDataset] was created
opt.netG:resnet_9blocks
initialize network with normal
model [AsyndganModel] was created
Namespace(aspect_ratio=1.0, batch_size=1, checkpoints_dir='./checkpoints', crop_size=256, dataroot='/home/OYSK/codes/AsynDGAN/test_demo', dataset_mode='brats', direction='AtoB', display_id=-1, display_winsize=256, epoch='latest', eval=False, gpu_ids=[0], init_gain=0.02, init_type='normal', input_nc=3, isTrain=False, load_iter=0, load_size=256, max_dataset_size=inf, model='asyndgan', n_layers_D=3, name='AtoB', ndf=64, netD='basic', netG='resnet_9blocks', ngf=64, no_dropout=False, no_flip=True, norm='batch', ntest=inf, num_test=inf, num_threads=0, output_nc=3, phase='test', preprocess='resize_and_crop', results_dir='results/test_asyndgan', serial_batches=True, suffix='', verbose=False)
loading the model from ./checkpoints/AtoB/latest_net_G.pth
---------- Networks initialized -------------
[Network G] Total number of parameters : 11.383 M
Traceback (most recent call last):
File "save_syn.py", line 115, in
model.set_input(data) # unpack data from data loader
File "/home/OYSK/codes/AsynDGAN/models/asyndgan_model.py", line 116, in set_input
self.real_A.append(input['A_' + str(i)].to(self.device))
KeyError: 'A_0'

python save_syn.py --dataroot /home/OYSK/codes/AsynDGAN/test --name AtoB --model asyndgan --netG resnet_9blocks --direction AtoB --dataset_mode brats_split --results_dir results/test_asyndgan
----------------- End -------------------
Load: BraTS18_tumor_size_0.h5
build key:0
Load: BraTS18_tumor_size_1.h5
build key:0
Load: BraTS18_tumor_size_2.h5
build key:0
dataset [BratsSplitDataset] was created
opt.netG:resnet_9blocks
initialize network with normal
model [AsyndganModel] was created
Namespace(aspect_ratio=1.0, batch_size=1, checkpoints_dir='./checkpoints', crop_size=256, dataroot='/home/OYSK/codes/AsynDGAN/test', dataset_mode='brats_split', direction='AtoB', display_id=-1, display_winsize=256, epoch='latest', eval=False, gpu_ids=[0], init_gain=0.02, init_type='normal', input_nc=3, isTrain=False, load_iter=0, load_size=256, max_dataset_size=inf, model='asyndgan', n_layers_D=3, name='AtoB', ndf=64, netD='basic', netG='resnet_9blocks', ngf=64, no_dropout=False, no_flip=True, norm='batch', ntest=inf, num_test=inf, num_threads=0, output_nc=3, phase='test', preprocess='resize_and_crop', results_dir='results/test_asyndgan', serial_batches=True, suffix='', verbose=False)
loading the model from ./checkpoints/AtoB/latest_net_G.pth
---------- Networks initialized -------------
[Network G] Total number of parameters : 11.383 M

Traceback (most recent call last):
File "save_syn.py", line 119, in
gt = data['Seg']
KeyError: 'Seg'

When testing, input a picture or label, and then generate another picture. I am weak, sorry. Hope you can give more guidance

from asyndgan.

Sorry for my late response. But it seems when you evaluating, you should directly use the pix2pix model rather than use asyndgan model since the evaluation stage should just use the central generator.

from asyndgan.