the tensor not mach

[Jarvis3310]

when i run
!sh ./scripts/test.sh
return this message, how could i fix?

+ GPU_ID=0
+ DISPLAY_ID=10
+ NAME=pifu_demo
+ BATCH_SIZE=1
+ MLP_DIM=257 1024 512 256 128 1
+ MLP_DIM_COLOR=513 1024 512 256 128 3
+ VOL_RES=256
+ CHECKPOINTS_NETG_PATH=./checkpoints/net_G
+ CHECKPOINTS_NETC_PATH=./checkpoints/net_C
+ TEST_FOLDER_PATH=./sample_images
+ CUDA_VISIBLE_DEVICES=0 python ./apps/eval.py --name pifu_demo --batch_size 1 --mlp_dim 257 1024 512 256 128 1 --mlp_dim_color 513 1024 512 256 128 3 --num_stack 4 --num_hourglass 2 --resolution 256 --hg_down ave_pool --norm group --norm_color group --test_folder_path ./sample_images --load_netG_checkpoint_path ./checkpoints/net_G --load_netC_checkpoint_path ./checkpoints/net_C
initialize network with normal
Using Network:  hgpifu
loading for net C ... ./checkpoints/net_C
initialize network with normal
num;  1
0it [00:00, ?it/s]./sample_images/abc.png ./sample_images/abc_mask.png
error: ('The size of tensor a (85) must match the size of tensor b (84) at non-singleton dimension 2',)
1it [00:06,  6.00s/it]

[Jarvis3310]

I solved it, the input image needs to be square, so I added this img1 = cv2.resize(img1, (512, 512), interpolation=cv2.INTER_AREA) to the generated blackwhite program