golang tensorflow client类型问题。。。

``` f, err := ioutil.ReadFile("2.npy") if err != nil { fmt.Println(err) return } tensor, err := tf.NewTensor(f) if err != nil { fmt.Println("Cannot read image file") return } tensorByte, ok := tensor.Value().([]byte) if !ok { fmt.Println("Cannot type assert tensor value to ") return } request := &pb.PredictRequest{ ModelSpec: &pb.ModelSpec{ Name: "wav", SignatureName: "voice_check", Version: &google_protobuf.Int64Value{ Value: int64(1), }, }, Inputs: map[string]*tf_core_framework.TensorProto{ "voice": &tf_core_framework.TensorProto{ Dtype: tf_core_framework.DataType_DT_FLOAT, TensorShape: &tf_core_framework.TensorShapeProto{ Dim: []*tf_core_framework.TensorShapeProto_Dim{ &tf_core_framework.TensorShapeProto_Dim{ Size: 1, }, &tf_core_framework.TensorShapeProto_Dim{ Size: 1, }, &tf_core_framework.TensorShapeProto_Dim{ Size: 96, }, &tf_core_framework.TensorShapeProto_Dim{ Size: 89, }, }, }, //TensorContent: f, StringVal: [][]byte{tensorByte}, }, }, } ``` 我使用python起了一个tensorflow的grpc服务端，想使用golang调用，但是需要传的类型是[][][][]float32，查了很多资料，没有找到相关，有大佬帮我解答一下吗。。。