Run pre-trained model for single image

Hi! I wanted to know if I can run the model on single image (either 224 x 224 or big tile(I'll resize)). 
I tried to do that but got error like **RuntimeError: Error(s) in loading state_dict for EncoderDecoder:
	size mismatch for decoder.decoder_pos_embed: copying a param with shape torch.Size([1, 289, 256]) from checkpoint, the shape in current model is torch.Size([1, 385, 256]).**