Tensorflow: load网络时出现 device='GPU' ...device='CPU...device= 'XLA_GPU'...等类型错误（import_meta_graph）

当你load tensorflow 的meta文件时

    new_saver = tf.compat.v1.train.import_meta_graph('path to meta')

出现如下的错误：

Op: MatMul
Node attrs: transpose_b=false, T=DT_INT32, transpose_a=false
Registered kernels:
  device='XLA_GPU'; T in [DT_FLOAT, DT_DOUBLE, DT_COMPLEX64, DT_BFLOAT16, DT_HALF]
  device='XLA_CPU'; T in [DT_FLOAT, DT_DOUBLE, DT_COMPLEX64, DT_BFLOAT16, DT_COMPLEX128, DT_HALF]
  device='XLA_CPU_JIT'; T in [DT_FLOAT, DT_DOUBLE, DT_COMPLEX64, DT_BFLOAT16, DT_COMPLEX128, DT_HALF]
  device='XLA_GPU_JIT'; T in [DT_FLOAT, DT_DOUBLE, DT_COMPLEX64, DT_BFLOAT16, DT_HALF]
  device='GPU'; label='cublas'; T in [DT_HALF]
  device='GPU'; label='cublas'; T in [DT_COMPLEX128]
  device='GPU'; label='cublas'; T in [DT_COMPLEX64]
  device='GPU'; label='cublas'; T in [DT_DOUBLE]
  device='GPU'; label='cublas'; T in [DT_FLOAT]
  device='GPU'; T in [DT_HALF]
  device='GPU'; T in [DT_COMPLEX128]
  device='GPU'; T in [DT_COMPLEX64]
  device='GPU'; T in [DT_DOUBLE]
  device='GPU'; T in [DT_FLOAT]
  device='CPU'; label='eigen'; T in [DT_COMPLEX128]
  device='CPU'; label='eigen'; T in [DT_COMPLEX64]
  device='CPU'; label='eigen'; T in [DT_INT64]
  device='CPU'; label='eigen'; T in [DT_INT32]
  device='CPU'; label='eigen'; T in [DT_BFLOAT16]
  device='CPU'; label='eigen'; T in [DT_HALF]
  device='CPU'; label='eigen'; T in [DT_DOUBLE]
  device='CPU'; label='eigen'; T in [DT_FLOAT]
  device='CPU'; T in [DT_COMPLEX128]
  device='CPU'; T in [DT_COMPLEX64]
  device='CPU'; T in [DT_INT64]
  device='CPU'; T in [DT_INT32]
  device='CPU'; T in [DT_BFLOAT16]
  device='CPU'; T in [DT_HALF]
  device='CPU'; T in [DT_DOUBLE]
  device='CPU'; T in [DT_FLOAT]

# 把clear_device设成True即可解决
new_saver = tf.compat.v1.train.import_meta_graph('path to meta', clear_devices=True)

问题解决！