InvalidArgumentError (see above for traceback): You must feed a value for placeholder tensor 'model_inputs/input_xc' with dtype float and shape [?,32,32,6]
[[node model_inputs/input_xc (defined at ..\models\STResNet.py:115) = Placeholder[dtype=DT_FLOAT, shape=[?,32,32,6], _device="/job:localhost/replica:0/task:0/device:CPU:0"]()]]
自我人为是展开的tensor的shape中有问号,例如shape(?,32,32,2).也不知道对不对,所以在显示张量的时候再往前看一看,找找有没有确定的shape。