DataLossError (see above for traceback): file is too short to be an sstable [[Node: save/RestoreV2 = RestoreV2[dtypes=[DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT, ..., DT_FLOAT, DT_FLOAT, DT_F

DataLossError (see above for traceback): file is too short to be an sstable
[[Node: save/RestoreV2 = RestoreV2[dtypes=[DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT, ..., DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_INT64], _device="/job:localhost/replica:0/task:0/device:CPU:0"](_arg_save/Const_0_0, save/RestoreV2/tensor_names, save/RestoreV2/shape_and_slices)]]

 一般是tensorflow 训练生成的model.ckpt文件有问题，我的这次就是文件里面没数据，于是我修改checkpoint里面的东西，删除掉没数据的model.ckpt，修改主model.ckpt，然后再训练或者推导就可以了