Thanks, I was able to load the model. Now I’m trying to work with it.

To find out input names I used `saved_model_cli`

and got this:

```
The given SavedModel SignatureDef contains the following input(s):
inputs['length'] tensor_info:
dtype: DT_INT32
shape: (-1)
name: Placeholder_1:0
inputs['tokens'] tensor_info:
dtype: DT_STRING
shape: (-1, -1)
name: Placeholder:0
The given SavedModel SignatureDef contains the following output(s):
outputs['length'] tensor_info:
dtype: DT_INT32
shape: (-1, 5)
name: seq2seq/decoder_1/decoder_1/while/Exit_16:0
outputs['log_probs'] tensor_info:
dtype: DT_FLOAT
shape: (-1, 5)
name: seq2seq/decoder_1/decoder_1/while/Exit_11:0
outputs['tokens'] tensor_info:
dtype: DT_STRING
shape: (-1, 5, -1)
name: seq2seq/index_to_string_Lookup:0
Method name is: tensorflow/serving/predict
```

So I tried to pass the input like so:

```
Session.Runner runner = model.session().runner();
String input = "random";
byte[][][] matrix = new byte[1][1][];
matrix[0][0] = input.getBytes("UTF-8");
Tensor<?> inputCallsTensor = Tensor.create(matrix, String.class);
Tensor<?> inputLengthTensor = Tensor.create(1);
runner.feed("Placeholder_1:0", inputLengthTensor);
runner.feed("Placeholder:0", inputCallsTensor);
List<Tensor<?>> run = runner.fetch("seq2seq/decoder_1/decoder_1/while/Exit_16:0").run();
```

This works well up to the last line. There an exception is thrown regarding wrong number of dimensions:

```
Exception in thread "main" java.lang.IllegalArgumentException: Tried to expand dim index 1 for tensor with 0 dimensions.
[[Node: seq2seq/decoder_1/tile_batch_2/ExpandDims = ExpandDims[T=DT_INT32, Tdim=DT_INT32, _output_shapes=[[?,1]], _device="/job:localhost/replica:0/task:0/device:CPU:0"](_arg_Placeholder_1_0_1, seq2seq/decoder_1/tile_batch/ExpandDims_2/dim)]]
at org.tensorflow.Session.run(Native Method)
```

I tried to make `matrix`

2-dimensional to no avail.

Do I understand correctly that input is a list of strings, and should be passed the way I pass it?