training directory structure / training with 2 images

[surfingnirvana]

Hi, i tried to train a model from scratch. When running pix2pix.py i get this error

(base) PS C:\NEURAL\Chinese-Landscape-Painting-Style-Transfer> python pix2pix.py
Namespace(epoch=0, n_epochs=200, dataset_name='alice', batch_size=2, lr=0.0002, b1=0.5, b2=0.999, decay_epoch=100, n_cpu=8, img_height=512, img_width=512, channels=3, sample_interval=3000, img_size=512, checkpoint_interval=5, output_path='pictures', texture_path='train_data/', latent_dim=256, n_critic=5, clip_value=0.01)
Traceback (most recent call last):
File "C:\NEURAL\Chinese-Landscape-Painting-Style-Transfer\pix2pix.py", line 67, in
dataset = FramesDataset(opt,dataset='alice',transform=transform)
File "C:\NEURAL\Chinese-Landscape-Painting-Style-Transfer\Frames_dataset.py", line 44, in init
assert(len(self.fnames)==len(self.fnames2))
AssertionError

I have these folders in my root:

--->train_data
------->Harvard
----------->canny&HED1.jpg
----------->canny&HED2.jpg

Where do i have to put the actual "alice" folder? with Alice-->Harvard-->img1 and so on...?
Is it possible to train with only two images?