some trouble in training

[limingxiang]

Hello everyone, when I used RGN to analyze the test data, I encountered the following error, when my server was running in centos7, python2 + tensorflow = = 1.4.0，test data is CASP7，no GPU device. the code is： python ~/rgn/model/protling.py ~/rgn/data/RGN7/runs/CASP7/ProteinNet7Thinning90/configuration -d ~/rgn/data/RGN7 Hope to get your help！

WARNING:tensorflow:Entity <bound method CudnnLSTM.call of <tensorflow.contrib.cudnn_rnn.python.layers.cudnn_rnn.CudnnLSTM object at 0x7fe776a66590>> could not be transformed and will be executed as-is. Please report this to the AutgoGraph team. When filing the bug, set the verbosity to 10 (on Linux, export AUTOGRAPH_VERBOSITY=10) and attach the full output. Cause: converting <bou nd method CudnnLSTM.call of <tensorflow.contrib.cudnn_rnn.python.layers.cudnn_rnn.CudnnLSTM object at 0x7fe776a66590>>: Attribut eError: 'module' object has no attribute 'Index' WARNING:tensorflow:Entity <bound method CudnnLSTM.call of <tensorflow.contrib.cudnn_rnn.python.layers.cudnn_rnn.CudnnLSTM object at 0x7fe776a667d0>> could not be transformed and will be executed as-is. Please report this to the AutgoGraph team. When filing the bug, set the verbosity to 10 (on Linux, export AUTOGRAPH_VERBOSITY=10) and attach the full output. Cause: converting <bou nd method CudnnLSTM.call of <tensorflow.contrib.cudnn_rnn.python.layers.cudnn_rnn.CudnnLSTM object at 0x7fe776a667d0>>: Attribut eError: 'module' object has no attribute 'Index' WARNING:tensorflow:Entity <bound method CudnnLSTM.call of <tensorflow.contrib.cudnn_rnn.python.layers.cudnn_rnn.CudnnLSTM object at 0x7fe776a80b90>> could not be transformed and will be executed as-is. Please report this to the AutgoGraph team. When filing ...skipping... nd method CudnnLSTM.call of <tensorflow.contrib.cudnn_rnn.python.layers.cudnn_rnn.CudnnLSTM object at 0x7fe776a66590>>: AttributeError: 'module' object has no attribute 'Index' WARNING:tensorflow:Entity <bound method CudnnLSTM.call of <tensorflow.contrib.cudnn_rnn.python.layers.cudnn_rnn.CudnnLSTM object at 0x7fe776a667d0>> could not be transformed and will be executed as-is. Please report this to the AutgoGraph team. When filing the bug, set the verbosity to 10 (on Linux, export AUTOGRAPH_VERBOSITY=10) and attach the full output. Cause: converting <bound method CudnnLSTM.call of <tensorflow.contrib.cudnn_rnn.python.layers.cudnn_rnn.CudnnLSTM object at 0x7fe776a667d0>>: AttributeError: 'module' object has no attribute 'Index' WARNING:tensorflow:Entity <bound method CudnnLSTM.call of <tensorflow.contrib.cudnn_rnn.python.layers.cudnn_rnn.CudnnLSTM object at 0x7fe776a80b90>> could not be transformed and will be executed as-is. Please report this to the AutgoGraph team. When filing the bug, set the verbosity to 10 (on Linux, export AUTOGRAPH_VERBOSITY=10) and attach the full output. Cause: converting <bound method CudnnLSTM.call of <tensorflow.contrib.cudnn_rnn.python.layers.cudnn_rnn.CudnnLSTM object at 0x7fe776a80b90>>: AttributeError: 'module' object has no attribute 'Index' WARNING:tensorflow:Entity <bound method CudnnLSTM.call of <tensorflow.contrib.cudnn_rnn.python.layers.cudnn_rnn.CudnnLSTM object at 0x7fe77684f810>> could not be transformed and will be executed as-is. Please report this to the AutgoGraph team. When filing the bug, set the verbosity to 10 (on Linux, export AUTOGRAPH_VERBOSITY=10) and attach the full output. Cause: converting <bound method CudnnLSTM.call of <tensorflow.contrib.cudnn_rnn.python.layers.cudnn_rnn.CudnnLSTM object at 0x7fe77684f810>>: AttributeError: 'module' object has no attribute 'Index' Traceback (most recent call last): File "/home/teng/rgn/model/protling.py", line 527, in while loop(args): pass File "/home/teng/rgn/model/protling.py", line 301, in loop models.update({'training': RGNModel('training', configs['training'])}) File "/home/teng/rgn/model/model.py", line 114, in init self._create_graph(mode, self.config) File "/home/teng/rgn/model/model.py", line 211, in _create_graph dihedrals = _dihedrals(mode, dihedrals_config, recurrent_outputs, alphabet=alphabet) File "/home/teng/rgn/model/model.py", line 950, in _dihedrals linear = layers.fully_connected(dihedrals_inputs, output_size, activation_fn=None, scope='linear_dihedrals', NameError: global name 'layers' is not defined