从tensorflow cpu切换到gpu时发生InvalidArgumentError

Question

我卸载了tensorflow cpu并安装了tensorflow-gpu 1.8.0，现在我所有的tensorflow或keras代码都给我这样的错误：

InvalidArgumentError: Input to reshape is a tensor with 4800 values, but the requested shape has 0
     [[Node: dense/Tensordot/Reshape = Reshape[T=DT_FLOAT, Tshape=DT_INT32, _device="/job:localhost/replica:0/task:0/device:GPU:0"](concat, dense/Tensordot/stack)]]
     [[Node: ArgMax/_59 = _Recv[client_terminated=false, recv_device="/job:localhost/replica:0/task:0/device:CPU:0", send_device="/job:localhost/replica:0/task:0/device:GPU:0", send_device_incarnation=1, tensor_name="edge_463_ArgMax", tensor_type=DT_INT64, _device="/job:localhost/replica:0/task:0/device:CPU:0"]()]]

在cpu上运行时100％正确。 我当前的安装是

• cudatoolkit 9.0
• cudnn 7.1.4
• tensorflow-gpu 1.8
• keras 2.1.2

我正在Windows 8.1上使用anaconda3。

感谢任何帮助。谢谢！