Dimension 0 in both shapes must be equal, but are 21 and 19. Shapes are [21] and [19]. for 'conv6_cls_2/Assign' (op: 'Assign') with input shapes: [21], [19].

[caijh23]

I cannot train with my own dataset using the pretrained model icnet_cityscapes_train_30k_bnnomerge.npy

[songshan0321]

How did u solve it? I get the same error even after doing this: restore_var = [v for v in tf.global_variables() if 'conv6_cls' not in v.name]