LSTM case test failed when process split op

I run the python script test_forward_ptb() in, and got the errors of
“TypeError: ‘Call’ object does not support indexing”

This error is caused by the split op, and the previous code was not appeared, Update the latest code that appears.Who has encountered similar problems, is there a solution?

2019-10-28 09:51:31.252852: I tensorflow/core/platform/] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA
Traceback (most recent call last):

  File "", line 2459, in <module>

  File "", line 1806, in test_forward_ptb
    params, m = _get_tvm_graph_module(graph_def)

  File "", line 1743, in _get_tvm_graph_module
    mod, params = relay.frontend.from_tensorflow(graph_def, shape=shape_dict)

  File "/home/Codes/tvm/python/tvm/relay/frontend/", line 2473, in from_tensorflow
    mod, params = g.from_tensorflow(graph, layout, shape, outputs, dtype)

  File "/home/Codes/tvm/python/tvm/relay/frontend/", line 2094, in from_tensorflow
    op = self._convert_operator(node.op, inputs, attr, graph)

  File "/home/Codes/tvm/python/tvm/relay/frontend/", line 2440, in _convert_operator

  File "/home/Codes/tvm/python/tvm/relay/frontend/", line 2317, in _convert_rnn_operator
    sym = self.rnn.process_op(op_name, inputs, attrs, params)

  File "/home/Codes/tvm/python/tvm/relay/frontend/", line 1634, in process_op
    params, num_layers)

  File "/home/Codes/tvm/python/tvm/relay/frontend/", line 1575, in _impl
    num_layers, self._cur_lstm_layer)

  File "/home/Codes/tvm/python/tvm/relay/frontend/", line 1567, in _LSTMBlockCellWrapper
    batch_size, num_hidden)

  File "/home/Codes/tvm/python/tvm/relay/frontend/", line 1541, in _get_cur_input_state
    cur_in_state_c = _op.reshape(in_state_c_tup[layer],

TypeError: 'Call' object does not support indexing