[ 2019-08-16 11:41:43,166 ] [ DEBUG ] [ main:134 ]  Namespace(batch=None, blobs_as_inputs=False, data_type='float', disable_fusing=None, disable_gfusing=None, disable_nhwc_to_nchw=False, disable_resnet_optimization=True, enable_concat_optimization=False, extensions='/opt/intel/openvino_2019.2.242/deployment_tools/model_optimizer/extensions', finegrain_fusing=None, framework='tf', freeze_placeholder_with_value=None, generate_deprecated_IR_V2=None, generate_experimental_IR_V10=False, input='inputs', input_checkpoint=None, input_meta_graph=None, input_model='/opt/intel/openvino_2019.2.242/deployment_tools/model_optimizer/model.pb', input_model_is_text=False, input_shape='[1,112,28,1]', keep_quantize_ops_in_IR=False, keep_shape_ops=False, log_level='DEBUG', mean_values=(), model_name=None, move_to_preprocess=False, output='seq_len', output_dir='/opt/intel/openvino_2019.2.242/deployment_tools/model_optimizer/.', reverse_input_channels=False, saved_model_dir=None, saved_model_tags=None, scale=None, scale_values=(), silent=False, tensorboard_logdir=None, tensorflow_custom_layer_libraries=None, tensorflow_custom_operations_config_update=None, tensorflow_object_detection_api_pipeline_config=None, tensorflow_operation_patterns=None, tensorflow_subgraph_patterns=None, tensorflow_use_custom_operations_config=None, version=False)
[ 2019-08-16 11:41:43,166 ] [ DEBUG ] [ main:135 ]  Model Optimizer started
[ 2019-08-16 11:41:43,166 ] [ DEBUG ] [ main:149 ]  Output model name would be model{.xml, .bin}
[ 2019-08-16 11:41:43,734 ] [ DEBUG ] [ main:235 ]  Placeholder shapes : {'inputs': array([  1, 112,  28,   1])}
[ INFO ]  Importing extensions from: /opt/intel/openvino_2019.2.242/deployment_tools/model_optimizer/mo
[ INFO ]  New subclass: <class 'extensions.ops.gather.Gather'>
[ INFO ]  Registered a new subclass with key: Gather
[ INFO ]  New subclass: <class 'mo.ops.const.Const'>
[ INFO ]  Registered a new subclass with key: Const
[ INFO ]  New subclass: <class 'extensions.ops.transpose.Transpose'>
[ INFO ]  Registered a new subclass with key: Transpose
[ INFO ]  New subclass: <class 'mo.ops.reshape.Reshape'>
[ INFO ]  Registered a new subclass with key: Reshape
[ INFO ]  New subclass: <class 'mo.ops.scale_shift.ScaleShiftOp'>
[ INFO ]  Registered a new subclass with key: ScaleShift
[ INFO ]  New subclass: <class 'mo.ops.clamp.Clamp'>
[ INFO ]  Registered a new subclass with key: Clamp
[ INFO ]  New subclass: <class 'extensions.ops.activation_ops.Activation'>
[ INFO ]  New subclass: <class 'extensions.ops.activation_ops.LeakyReLU'>
[ INFO ]  Registered a new subclass with key: LeakyReLU
[ INFO ]  New subclass: <class 'mo.ops.concat.Concat'>
[ INFO ]  Registered a new subclass with key: Concat
[ INFO ]  New subclass: <class 'mo.ops.convolution.Convolution'>
[ INFO ]  Registered a new subclass with key: Convolution
[ INFO ]  New subclass: <class 'mo.ops.crop.Crop'>
[ INFO ]  Registered a new subclass with key: Crop
[ INFO ]  New subclass: <class 'mo.ops.softmax.Softmax'>
[ INFO ]  Registered a new subclass with key: SoftMax
[ INFO ]  New subclass: <class 'extensions.ops.elementwise.Elementwise'>
[ INFO ]  New subclass: <class 'mo.ops.activation.Activation'>
[ INFO ]  Registered a new subclass with key: Activation
[ INFO ]  New subclass: <class 'mo.ops.broadcast.Broadcast'>
[ INFO ]  Registered a new subclass with key: Broadcast
[ INFO ]  New subclass: <class 'mo.ops.deconvolution.Deconvolution'>
[ INFO ]  Registered a new subclass with key: Deconvolution
[ INFO ]  New subclass: <class 'mo.ops.deformable_convolution.DeformableConvolution'>
[ INFO ]  Registered a new subclass with key: DeformableConvolution
[ INFO ]  New subclass: <class 'mo.ops.eltwise.Eltwise'>
[ INFO ]  Registered a new subclass with key: Eltwise
[ INFO ]  New subclass: <class 'mo.ops.eltwise_n.EltwiseN'>
[ INFO ]  Registered a new subclass with key: EltwiseN
[ INFO ]  New subclass: <class 'mo.ops.eltwise_ninputs_in_1.EltwiseNin1'>
[ INFO ]  Registered a new subclass with key: EltwiseNin1
[ INFO ]  New subclass: <class 'mo.ops.expand_dims.ExpandDims'>
[ INFO ]  Registered a new subclass with key: ExpandDims
[ INFO ]  New subclass: <class 'mo.ops.fill.Fill'>
[ INFO ]  Registered a new subclass with key: Fill
[ INFO ]  New subclass: <class 'mo.ops.flatten.Flatten'>
[ INFO ]  Registered a new subclass with key: Flatten
[ INFO ]  New subclass: <class 'mo.ops.flatten_onnx.FlattenONNX'>
[ INFO ]  Registered a new subclass with key: FlattenONNX
[ INFO ]  New subclass: <class 'mo.ops.group_norm.GroupNorm'>
[ INFO ]  Registered a new subclass with key: GroupNorm
[ INFO ]  New subclass: <class 'mo.ops.inner_product.InnerProduct'>
[ INFO ]  Registered a new subclass with key: FullyConnected
[ INFO ]  New subclass: <class 'mo.ops.lrn.LRN'>
[ INFO ]  Registered a new subclass with key: LRN
[ INFO ]  New subclass: <class 'mo.ops.lstmnonlinearity.LstmNonLinearity'>
[ INFO ]  Registered a new subclass with key: LstmNonLinearity
[ INFO ]  New subclass: <class 'mo.ops.memory.Memory'>
[ INFO ]  Registered a new subclass with key: Memory
[ INFO ]  New subclass: <class 'mo.ops.memoryoffset.MemoryOffset'>
[ INFO ]  Registered a new subclass with key: MemoryOffset
[ INFO ]  New subclass: <class 'mo.ops.pad.Pad'>
[ INFO ]  Registered a new subclass with key: Pad
[ INFO ]  New subclass: <class 'mo.ops.permute.Permute'>
[ INFO ]  Registered a new subclass with key: Permute
[ INFO ]  New subclass: <class 'mo.ops.pooling.Pooling'>
[ INFO ]  Registered a new subclass with key: Pooling
[ INFO ]  New subclass: <class 'mo.ops.power.Power'>
[ INFO ]  Registered a new subclass with key: Power
[ INFO ]  New subclass: <class 'mo.ops.result.Result'>
[ INFO ]  Registered a new subclass with key: Result
[ INFO ]  New subclass: <class 'mo.ops.roipooling.ROIPooling'>
[ INFO ]  Registered a new subclass with key: ROIPooling
[ INFO ]  New subclass: <class 'mo.ops.shape.Shape'>
[ INFO ]  Registered a new subclass with key: ShapeOf
[ INFO ]  New subclass: <class 'mo.ops.slice.Slice'>
[ INFO ]  Registered a new subclass with key: Slice
[ INFO ]  New subclass: <class 'mo.ops.split.Split'>
[ INFO ]  Registered a new subclass with key: Split
[ INFO ]  New subclass: <class 'mo.ops.squeeze.Squeeze'>
[ INFO ]  Registered a new subclass with key: Squeeze
[ INFO ]  New subclass: <class 'mo.ops.strided_slice.StridedSlice'>
[ INFO ]  Registered a new subclass with key: StridedSlice
[ INFO ]  New subclass: <class 'mo.ops.tile.Tile'>
[ INFO ]  Registered a new subclass with key: Tile
[ INFO ]  New subclass: <class 'mo.ops.unsqueeze.Unsqueeze'>
[ INFO ]  Registered a new subclass with key: Unsqueeze
[ INFO ]  New subclass: <class 'mo.front.common.replacement.FrontReplacementSubgraph'>
[ INFO ]  New subclass: <class 'mo.front.tf.replacement.FrontReplacementFromConfigFileGeneral'>
[ INFO ]  New subclass: <class 'mo.front.common.replacement.FrontReplacementOp'>
[ INFO ]  Registered a new subclass with key: UnknownOp
[ INFO ]  New subclass: <class 'mo.front.tf.replacement.FrontReplacementFromConfigFileSubGraph'>
[ INFO ]  New subclass: <class 'mo.front.tf.replacement.FrontReplacementFromConfigFileOp'>
[ WARNING ]  Skipped <class 'mo.front.common.replacement.FrontReplacementSubgraph'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.front.tf.replacement.FrontReplacementFromConfigFileGeneral'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.front.common.replacement.FrontReplacementOp'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.front.tf.replacement.FrontReplacementFromConfigFileSubGraph'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.front.tf.replacement.FrontReplacementFromConfigFileOp'> registration because it was already registered or it was disabled. 
[ INFO ]  New subclass: <class 'extensions.middle.ConcatOptimization.ConcatOptimization'>
[ INFO ]  New subclass: <class 'extensions.middle.EltwiseChecker.EltwiseChecker'>
[ INFO ]  New subclass: <class 'extensions.middle.EltwiseInputReshape.Eltwise1DInputReshape'>
[ INFO ]  New subclass: <class 'extensions.middle.EltwiseInputReshape.EltwiseInputReshape'>
[ INFO ]  New subclass: <class 'extensions.middle.EltwiseInputNormalization.EltwiseInputNormalize'>
[ INFO ]  New subclass: <class 'extensions.back.CreateConstNodes.CreateConstNodesReplacement'>
[ INFO ]  New subclass: <class 'extensions.back.FuseReshapesSequence.FuseReshapesSequence'>
[ INFO ]  New subclass: <class 'extensions.back.InsertLayoutPropagationTransposes.InsertLayoutPropagationTranspose'>
[ INFO ]  New subclass: <class 'extensions.back.RemoveRedundantReshapes.RemoveRedundantReshapes'>
[ INFO ]  New subclass: <class 'extensions.middle.LayoutChangeForConstantShapePaths.LayoutChangeForConstantShapePaths'>
[ INFO ]  New subclass: <class 'mo.front.tf.extractors.strided_slice.StridedSliceFrontExtractor'>
[ INFO ]  Registered a new subclass with key: StridedSlice
[ WARNING ]  Skipped <class 'mo.front.common.replacement.FrontReplacementSubgraph'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.front.tf.replacement.FrontReplacementFromConfigFileGeneral'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.front.common.replacement.FrontReplacementOp'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.front.tf.replacement.FrontReplacementFromConfigFileSubGraph'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.front.tf.replacement.FrontReplacementFromConfigFileOp'> registration because it was already registered or it was disabled. 
[ INFO ]  Importing extensions from: /opt/intel/openvino_2019.2.242/deployment_tools/model_optimizer/extensions
[ WARNING ]  Skipped <class 'extensions.ops.gather.Gather'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.const.Const'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.ops.transpose.Transpose'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.reshape.Reshape'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.scale_shift.ScaleShiftOp'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.clamp.Clamp'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.ops.activation_ops.Activation'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.ops.activation_ops.LeakyReLU'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.concat.Concat'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.convolution.Convolution'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.crop.Crop'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.softmax.Softmax'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.ops.elementwise.Elementwise'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.activation.Activation'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.broadcast.Broadcast'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.deconvolution.Deconvolution'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.deformable_convolution.DeformableConvolution'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.eltwise.Eltwise'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.eltwise_n.EltwiseN'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.eltwise_ninputs_in_1.EltwiseNin1'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.expand_dims.ExpandDims'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.fill.Fill'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.flatten.Flatten'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.flatten_onnx.FlattenONNX'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.group_norm.GroupNorm'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.inner_product.InnerProduct'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.lrn.LRN'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.lstmnonlinearity.LstmNonLinearity'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.memory.Memory'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.memoryoffset.MemoryOffset'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.pad.Pad'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.permute.Permute'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.pooling.Pooling'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.power.Power'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.result.Result'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.roipooling.ROIPooling'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.shape.Shape'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.slice.Slice'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.split.Split'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.squeeze.Squeeze'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.strided_slice.StridedSlice'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.tile.Tile'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.ops.unsqueeze.Unsqueeze'> registration because it was already registered or it was disabled. 
[ INFO ]  New subclass: <class 'extensions.ops.BatchMatMul.BatchMatMul'>
[ INFO ]  Registered a new subclass with key: BatchMatMul
[ INFO ]  New subclass: <class 'extensions.ops.BlockLSTM.BlockLSTM'>
[ INFO ]  Registered a new subclass with key: BlockLSTM
[ INFO ]  New subclass: <class 'extensions.ops.Cast.Cast'>
[ INFO ]  Registered a new subclass with key: Cast
[ INFO ]  New subclass: <class 'extensions.ops.DetectionOutput.DetectionOutput'>
[ INFO ]  Registered a new subclass with key: DetectionOutput
[ INFO ]  New subclass: <class 'extensions.ops.Enter.Enter'>
[ INFO ]  Registered a new subclass with key: Enter
[ INFO ]  New subclass: <class 'extensions.ops.Exit.Exit'>
[ INFO ]  Registered a new subclass with key: Exit
[ INFO ]  New subclass: <class 'extensions.ops.RNN.RNN'>
[ INFO ]  Registered a new subclass with key: RNN
[ INFO ]  New subclass: <class 'extensions.ops.GRU.GRU'>
[ INFO ]  Registered a new subclass with key: GRU
[ INFO ]  New subclass: <class 'extensions.ops.GRUCell.GRUCell'>
[ INFO ]  Registered a new subclass with key: GRUCell
[ INFO ]  New subclass: <class 'extensions.ops.GatherNd.GatherNd'>
[ INFO ]  Registered a new subclass with key: GatherNd
[ INFO ]  New subclass: <class 'extensions.ops.GatherTree.GatherTree'>
[ INFO ]  Registered a new subclass with key: GatherTree
[ INFO ]  New subclass: <class 'extensions.ops.LSTM.LSTM'>
[ INFO ]  Registered a new subclass with key: LSTM
[ INFO ]  New subclass: <class 'extensions.ops.Log.LogOp'>
[ INFO ]  Registered a new subclass with key: Log
[ INFO ]  New subclass: <class 'extensions.ops.NextIteration.NextIteration'>
[ INFO ]  Registered a new subclass with key: NextIteration
[ INFO ]  New subclass: <class 'extensions.ops.RNNCell.RNNCell'>
[ INFO ]  Registered a new subclass with key: RNNCell
[ INFO ]  New subclass: <class 'extensions.ops.ReduceOps.ReduceOp'>
[ INFO ]  New subclass: <class 'extensions.ops.Reverse.Reverse'>
[ INFO ]  Registered a new subclass with key: Reverse
[ INFO ]  New subclass: <class 'extensions.ops.SquaredDifference.SquaredDifference'>
[ INFO ]  Registered a new subclass with key: SquaredDifference
[ INFO ]  New subclass: <class 'extensions.ops.TensorArray.TensorArray'>
[ INFO ]  Registered a new subclass with key: TensorArrayV3
[ INFO ]  New subclass: <class 'extensions.ops.TensorArrayGather.TensorArrayGather'>
[ INFO ]  Registered a new subclass with key: TensorArrayGatherV3
[ INFO ]  New subclass: <class 'extensions.ops.TensorArrayRead.TensorArrayReader'>
[ INFO ]  Registered a new subclass with key: TensorArrayReadV3
[ INFO ]  New subclass: <class 'extensions.ops.TensorArrayScatter.TensorArrayScatter'>
[ INFO ]  Registered a new subclass with key: TensorArrayScatterV3
[ INFO ]  New subclass: <class 'extensions.ops.TensorArraySize.TensorArraySize'>
[ INFO ]  Registered a new subclass with key: TensorArraySizeV3
[ INFO ]  New subclass: <class 'extensions.ops.TensorArrayWrite.TensorArrayWriter'>
[ INFO ]  Registered a new subclass with key: TensorArrayWriteV3
[ INFO ]  New subclass: <class 'extensions.ops.TensorIterator_ops.TensorIteratorInput'>
[ INFO ]  Registered a new subclass with key: TensorIteratorInput
[ INFO ]  New subclass: <class 'extensions.ops.TensorIterator_ops.TensorIteratorOutput'>
[ INFO ]  Registered a new subclass with key: TensorIteratorOutput
[ INFO ]  New subclass: <class 'extensions.ops.TensorIterator_ops.TensorIteratorCondition'>
[ INFO ]  Registered a new subclass with key: TensorIteratorCondition
[ INFO ]  New subclass: <class 'extensions.ops.TensorIterator_ops.TensorIteratorBackEdge'>
[ INFO ]  Registered a new subclass with key: TensorIteratorBackEdge
[ INFO ]  New subclass: <class 'extensions.ops.accum.AccumOp'>
[ INFO ]  Registered a new subclass with key: Accum
[ INFO ]  New subclass: <class 'extensions.ops.argmax.ArgMaxOp'>
[ INFO ]  Registered a new subclass with key: ArgMax
[ INFO ]  New subclass: <class 'extensions.ops.assert_op.Assert'>
[ INFO ]  Registered a new subclass with key: Assert
[ INFO ]  New subclass: <class 'extensions.ops.axpy.AxpyOp'>
[ INFO ]  Registered a new subclass with key: Axpy
[ INFO ]  New subclass: <class 'extensions.ops.binarization.Binarization'>
[ INFO ]  Registered a new subclass with key: Binarization
[ INFO ]  New subclass: <class 'extensions.ops.bn.BNOp'>
[ INFO ]  Registered a new subclass with key: BN
[ INFO ]  New subclass: <class 'extensions.ops.box_nms.BoxNms'>
[ INFO ]  Registered a new subclass with key: _contrib_box_nms
[ INFO ]  New subclass: <class 'extensions.ops.constant_fill.ConstantFill'>
[ INFO ]  Registered a new subclass with key: ConstantFill
[ INFO ]  New subclass: <class 'extensions.ops.copy.CopyOp'>
[ INFO ]  Registered a new subclass with key: copy
[ INFO ]  New subclass: <class 'extensions.ops.correlation.CorrelationOp'>
[ INFO ]  Registered a new subclass with key: Correlation
[ INFO ]  New subclass: <class 'extensions.ops.ctc_greedy_decoder.CTCGreedyDecoderOp'>
[ INFO ]  Registered a new subclass with key: CTCGreedyDecoder
[ INFO ]  New subclass: <class 'extensions.ops.data_augmentation.DataAugmentationOp'>
[ INFO ]  Registered a new subclass with key: DataAugmentation
[ INFO ]  New subclass: <class 'extensions.ops.depth_to_space.DepthToSpaceOp'>
[ INFO ]  Registered a new subclass with key: DepthToSpace
[ INFO ]  New subclass: <class 'extensions.ops.detectionoutput_onnx.ExperimentalDetectronDetectionOutput'>
[ INFO ]  Registered a new subclass with key: ExperimentalDetectronDetectionOutput
[ INFO ]  New subclass: <class 'extensions.ops.exp.ExpOp'>
[ INFO ]  Registered a new subclass with key: Exp
[ INFO ]  New subclass: <class 'extensions.ops.fakequantize.FakeQuantize'>
[ INFO ]  Registered a new subclass with key: FakeQuantize
[ INFO ]  New subclass: <class 'extensions.ops.grn.GRNOp'>
[ INFO ]  Registered a new subclass with key: GRN
[ INFO ]  New subclass: <class 'extensions.ops.identity.IdentityOp'>
[ INFO ]  Registered a new subclass with key: Identity
[ INFO ]  New subclass: <class 'extensions.ops.instance_normalization.InstanceNormalization'>
[ INFO ]  Registered a new subclass with key: InstanceNormalization
[ INFO ]  New subclass: <class 'extensions.ops.interp.InterpOp'>
[ INFO ]  Registered a new subclass with key: Interp
[ INFO ]  New subclass: <class 'extensions.ops.interpolate.Interpolate'>
[ INFO ]  Registered a new subclass with key: Interpolate
[ INFO ]  New subclass: <class 'extensions.ops.lstm_cell.LSTMCell'>
[ INFO ]  Registered a new subclass with key: LSTMCell
[ INFO ]  New subclass: <class 'extensions.ops.lstm_sequence.LSTMSequence'>
[ INFO ]  Registered a new subclass with key: LSTMSequence
[ INFO ]  New subclass: <class 'extensions.ops.merge.Merge'>
[ INFO ]  Registered a new subclass with key: Merge
[ INFO ]  New subclass: <class 'extensions.ops.mvn.MVN'>
[ INFO ]  Registered a new subclass with key: MVN
[ INFO ]  New subclass: <class 'extensions.ops.normalize.NormalizeOp'>
[ INFO ]  Registered a new subclass with key: Normalize
[ INFO ]  New subclass: <class 'extensions.ops.one_hot.OneHot'>
[ INFO ]  Registered a new subclass with key: OneHot
[ INFO ]  New subclass: <class 'extensions.ops.pack.PackOp'>
[ INFO ]  Registered a new subclass with key: Pack
[ INFO ]  New subclass: <class 'extensions.ops.parameter.Parameter'>
[ INFO ]  Registered a new subclass with key: Parameter
[ INFO ]  New subclass: <class 'extensions.ops.power_file.PowerFileOp'>
[ INFO ]  Registered a new subclass with key: PowerFile
[ INFO ]  New subclass: <class 'extensions.ops.prediction_heatmap.PredictionHeatmapOp'>
[ INFO ]  Registered a new subclass with key: PredictionHeatmap
[ INFO ]  New subclass: <class 'extensions.ops.prelu.PreluOp'>
[ INFO ]  Registered a new subclass with key: PReLU
[ INFO ]  New subclass: <class 'extensions.ops.priorbox.PriorBoxOp'>
[ INFO ]  Registered a new subclass with key: PriorBox
[ INFO ]  New subclass: <class 'extensions.ops.priorbox_clustered.PriorBoxClusteredOp'>
[ INFO ]  Registered a new subclass with key: PriorBoxClustered
[ INFO ]  New subclass: <class 'extensions.ops.priorgridgenerator_onnx.ExperimentalDetectronPriorGridGenerator'>
[ INFO ]  Registered a new subclass with key: ExperimentalDetectronPriorGridGenerator
[ INFO ]  New subclass: <class 'extensions.ops.proposal.ProposalOp'>
[ INFO ]  Registered a new subclass with key: Proposal
[ INFO ]  New subclass: <class 'extensions.ops.proposal_onnx.ExperimentalDetectronGenerateProposalsSingleImage'>
[ INFO ]  Registered a new subclass with key: ExperimentalDetectronGenerateProposalsSingleImage
[ INFO ]  New subclass: <class 'extensions.ops.psroipooling.PSROIPoolingOp'>
[ INFO ]  Registered a new subclass with key: PSROIPooling
[ INFO ]  New subclass: <class 'extensions.ops.range.Range'>
[ INFO ]  Registered a new subclass with key: Range
[ INFO ]  New subclass: <class 'extensions.ops.rank.Rank'>
[ INFO ]  Registered a new subclass with key: Rank
[ INFO ]  New subclass: <class 'extensions.ops.regionyolo.RegionYoloOp'>
[ INFO ]  Registered a new subclass with key: RegionYolo
[ INFO ]  New subclass: <class 'extensions.ops.reorgyolo.ReorgYoloOp'>
[ INFO ]  Registered a new subclass with key: ReorgYolo
[ INFO ]  New subclass: <class 'extensions.ops.resample.ResampleOp'>
[ INFO ]  Registered a new subclass with key: Resample
[ INFO ]  New subclass: <class 'extensions.ops.reverse_sequence.ReverseSequence'>
[ INFO ]  Registered a new subclass with key: ReverseSequence
[ INFO ]  New subclass: <class 'extensions.ops.roifeatureextractor_onnx.ExperimentalDetectronROIFeatureExtractor'>
[ INFO ]  Registered a new subclass with key: ExperimentalDetectronROIFeatureExtractor
[ INFO ]  New subclass: <class 'extensions.ops.select.Select'>
[ INFO ]  Registered a new subclass with key: Select
[ INFO ]  New subclass: <class 'extensions.ops.shufflechannel.ShuffleChannelOp'>
[ INFO ]  Registered a new subclass with key: ShuffleChannel
[ INFO ]  New subclass: <class 'extensions.ops.simplernms.SimplerNMSOp'>
[ INFO ]  Registered a new subclass with key: SimplerNMS
[ INFO ]  New subclass: <class 'extensions.ops.size.Size'>
[ INFO ]  Registered a new subclass with key: Size
[ INFO ]  New subclass: <class 'extensions.ops.space_to_depth.SpaceToDepth'>
[ INFO ]  Registered a new subclass with key: SpaceToDepth
[ INFO ]  New subclass: <class 'extensions.ops.spatial_transformer.SpatialTransformOp'>
[ INFO ]  Registered a new subclass with key: SpatialTransformer
[ INFO ]  New subclass: <class 'extensions.ops.splice.Splice'>
[ INFO ]  Registered a new subclass with key: Splice
[ INFO ]  New subclass: <class 'extensions.ops.splitv.SplitV'>
[ INFO ]  Registered a new subclass with key: SplitV
[ INFO ]  New subclass: <class 'extensions.ops.stop_gradient.StopGradientOp'>
[ INFO ]  Registered a new subclass with key: StopGradient
[ INFO ]  New subclass: <class 'extensions.ops.swapaxis.SwapAxis'>
[ INFO ]  Registered a new subclass with key: SwapAxis
[ INFO ]  New subclass: <class 'extensions.ops.switch.Switch'>
[ INFO ]  Registered a new subclass with key: Switch
[ INFO ]  New subclass: <class 'extensions.ops.tensor_iterator.TensorIterator'>
[ INFO ]  Registered a new subclass with key: TensorIterator
[ INFO ]  New subclass: <class 'extensions.ops.topk.TopK'>
[ INFO ]  Registered a new subclass with key: TopK
[ INFO ]  New subclass: <class 'extensions.ops.topkrois_onnx.ExperimentalDetectronTopKROIs'>
[ INFO ]  Registered a new subclass with key: ExperimentalDetectronTopKROIs
[ INFO ]  New subclass: <class 'extensions.ops.upsample.UpsampleOp'>
[ INFO ]  Registered a new subclass with key: Upsample
[ WARNING ]  Skipped <class 'mo.front.tf.extractors.strided_slice.StridedSliceFrontExtractor'> registration because it was already registered or it was disabled. 
[ INFO ]  New subclass: <class 'extensions.front.InterpolateNormalizer.InterpolateNormalizer'>
[ INFO ]  Registered a new subclass with key: Interpolate
[ INFO ]  New subclass: <class 'extensions.front.LRNReplacer.LRNReplacer'>
[ INFO ]  Registered a new subclass with key: LRN
[ INFO ]  New subclass: <class 'extensions.front.Pack.Pack'>
[ INFO ]  Registered a new subclass with key: Pack
[ INFO ]  New subclass: <class 'extensions.front.PowerToEltwises.PowerToEltwises'>
[ INFO ]  Registered a new subclass with key: Power
[ INFO ]  New subclass: <class 'extensions.front.tf.FakeQuantWithMinMaxVars.FakeQuantWithMinMaxVarsToQuantize'>
[ INFO ]  Registered a new subclass with key: FakeQuantWithMinMaxVars
[ INFO ]  New subclass: <class 'extensions.front.div.Div'>
[ INFO ]  Registered a new subclass with key: Div
[ INFO ]  New subclass: <class 'extensions.front.eltwise_n.EltwiseNReplacement'>
[ INFO ]  Registered a new subclass with key: EltwiseN
[ INFO ]  New subclass: <class 'extensions.front.image_scaler.ImageScaler'>
[ INFO ]  Registered a new subclass with key: ImageScaler
[ INFO ]  New subclass: <class 'extensions.front.instance_normalization.InstanceNormalization'>
[ INFO ]  Registered a new subclass with key: InstanceNormalization
[ INFO ]  New subclass: <class 'extensions.front.reciprocal.ReciprocalReplacer'>
[ INFO ]  Registered a new subclass with key: Reciprocal
[ INFO ]  New subclass: <class 'extensions.front.sub.Sub'>
[ INFO ]  Registered a new subclass with key: Sub
[ INFO ]  New subclass: <class 'extensions.front.squared_difference.SquaredDifference'>
[ INFO ]  Registered a new subclass with key: SquaredDifference
[ WARNING ]  Skipped <class 'mo.front.common.replacement.FrontReplacementSubgraph'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.front.tf.replacement.FrontReplacementFromConfigFileGeneral'> registration because it was already registered or it was disabled. 
[ INFO ]  New subclass: <class 'extensions.front.ExpandDimsToUnsqueeze.ExpandDimsToUnsqueeze'>
[ INFO ]  New subclass: <class 'extensions.front.FillToBroadcast.FillToBroadcast'>
[ INFO ]  New subclass: <class 'extensions.front.SqueezeNormalize.SqueezeNormalize'>
[ INFO ]  New subclass: <class 'extensions.front.TopKNormalize.TopKNormalize'>
[ INFO ]  New subclass: <class 'extensions.front.binary_quantize_normalization.BinaryFakeQuantizeNormalization'>
[ INFO ]  New subclass: <class 'extensions.front.create_tensor_nodes.CreateTensorNodes'>
[ INFO ]  New subclass: <class 'extensions.front.disable_weights_quantize_value_propagation.DisableQuantizeValuePropagation'>
[ INFO ]  New subclass: <class 'extensions.front.input_cut.InputCut'>
[ INFO ]  New subclass: <class 'extensions.front.output_cut.OutputCut'>
[ INFO ]  New subclass: <class 'extensions.front.override_batch.OverrideBatch'>
[ INFO ]  New subclass: <class 'extensions.front.pass_separator.FrontStart'>
[ INFO ]  New subclass: <class 'extensions.front.pass_separator.FrontFinish'>
[ INFO ]  New subclass: <class 'extensions.front.user_data_repack.UserDataRepack'>
[ WARNING ]  Skipped <class 'mo.front.common.replacement.FrontReplacementOp'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.front.tf.replacement.FrontReplacementFromConfigFileSubGraph'> registration because it was already registered or it was disabled. 
[ INFO ]  New subclass: <class 'extensions.front.ArgMaxSqueeze.ArgMaxSqueeze'>
[ INFO ]  Registered a new subclass with key: ArgMax
[ INFO ]  New subclass: <class 'extensions.front.TransposeOrderNormalizer.TransposeOrderNormalizer'>
[ INFO ]  New subclass: <class 'extensions.front.freeze_placeholder_value.FreezePlaceholderValue'>
[ INFO ]  New subclass: <class 'extensions.front.no_op_eraser.NoOpEraser'>
[ INFO ]  New subclass: <class 'extensions.front.reduce_axis_normalizer.ReduceAxisNormalizer'>
[ INFO ]  New subclass: <class 'extensions.front.reshape_dim_normalizer.ReshapeDimNormalizer'>
[ INFO ]  New subclass: <class 'extensions.front.restore_ports.RestorePorts'>
[ INFO ]  New subclass: <class 'extensions.front.softmax.SoftmaxFromKeras'>
[ INFO ]  New subclass: <class 'extensions.front.standalone_const_eraser.StandaloneConstEraser'>
[ WARNING ]  Skipped <class 'mo.front.tf.replacement.FrontReplacementFromConfigFileOp'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.front.tf.extractors.strided_slice.StridedSliceFrontExtractor'> registration because it was already registered or it was disabled. 
[ INFO ]  New subclass: <class 'extensions.front.tf.BatchMatMul_ext.BatchMatMulExtractor'>
[ INFO ]  Registered a new subclass with key: BatchMatMul
[ INFO ]  New subclass: <class 'extensions.front.tf.BlockLSTM_ext.BlockLSTMExtractor'>
[ INFO ]  Registered a new subclass with key: BlockLSTM
[ INFO ]  New subclass: <class 'extensions.front.tf.CTCGreedyDecoder_ext.CTCCGreedyDecoderFrontExtractor'>
[ INFO ]  Registered a new subclass with key: CTCGreedyDecoder
[ INFO ]  New subclass: <class 'extensions.front.tf.Cast_ext.CastFrontExtractor'>
[ INFO ]  Registered a new subclass with key: Cast
[ INFO ]  New subclass: <class 'extensions.front.tf.FakeQuantWithMinMaxVars_ext.FakeQuantWithMinMaxVarsExtractor'>
[ INFO ]  Registered a new subclass with key: FakeQuantWithMinMaxVars
[ INFO ]  New subclass: <class 'extensions.front.tf.FloorMod_ext.EltwiseFrontExtractor'>
[ INFO ]  Registered a new subclass with key: FloorMod
[ INFO ]  New subclass: <class 'extensions.front.tf.GatherTree_ext.GatherTreeFrontExtractor'>
[ INFO ]  Registered a new subclass with key: GatherTree
[ INFO ]  New subclass: <class 'extensions.front.tf.LoopCond_ext.LoopCondFrontExtractor'>
[ INFO ]  Registered a new subclass with key: LoopCond
[ INFO ]  New subclass: <class 'extensions.front.tf.Not_ext.LogicalNotFrontExtractor'>
[ INFO ]  Registered a new subclass with key: LogicalNot
[ INFO ]  New subclass: <class 'extensions.front.tf.TensorArrayExtractors.TensorArrayExtractor'>
[ INFO ]  Registered a new subclass with key: TensorArrayV3
[ INFO ]  New subclass: <class 'extensions.front.tf.TensorArrayGatherV3.TensorArrayGatherV3Exteractor'>
[ INFO ]  Registered a new subclass with key: TensorArrayGatherV3
[ INFO ]  New subclass: <class 'extensions.front.tf.all_ext.AllFrontExtractor'>
[ INFO ]  Registered a new subclass with key: All
[ INFO ]  New subclass: <class 'extensions.front.tf.argmax_ext.ArgMaxFrontExtractor'>
[ INFO ]  Registered a new subclass with key: ArgMax
[ INFO ]  New subclass: <class 'extensions.front.tf.concat_ext.ConcatFrontExtractor'>
[ INFO ]  Registered a new subclass with key: Concat
[ INFO ]  New subclass: <class 'extensions.front.tf.conv_ext.Conv2DFrontExtractor'>
[ INFO ]  Registered a new subclass with key: Conv2D
[ INFO ]  New subclass: <class 'extensions.front.tf.conv_ext.DepthwiseConv2dNativeFrontExtractor'>
[ INFO ]  Registered a new subclass with key: DepthwiseConv2dNative
[ INFO ]  New subclass: <class 'extensions.front.tf.conv_ext.Conv3DFrontExtractor'>
[ INFO ]  Registered a new subclass with key: Conv3D
[ INFO ]  New subclass: <class 'extensions.front.tf.crop_and_resize_ext.CropAndResizeFrontExtractor'>
[ INFO ]  Registered a new subclass with key: CropAndResize
[ INFO ]  New subclass: <class 'extensions.front.tf.deconv_ext.Conv2DBackpropInputFrontExtractor'>
[ INFO ]  Registered a new subclass with key: Conv2DBackpropInput
[ INFO ]  New subclass: <class 'extensions.front.tf.deconv_ext.Conv3DBackpropInputV2InputFrontExtractor'>
[ INFO ]  Registered a new subclass with key: Conv3DBackpropInputV2
[ INFO ]  New subclass: <class 'extensions.front.tf.depth_to_space.DepthToSpaceFrontExtractor'>
[ INFO ]  Registered a new subclass with key: DepthToSpace
[ INFO ]  New subclass: <class 'extensions.front.tf.elementwise_ext.AddExtractor'>
[ INFO ]  Registered a new subclass with key: Add
[ INFO ]  New subclass: <class 'extensions.front.tf.elementwise_ext.AddNExtractor'>
[ INFO ]  Registered a new subclass with key: AddN
[ INFO ]  New subclass: <class 'extensions.front.tf.elementwise_ext.BiasAddExtractor'>
[ INFO ]  Registered a new subclass with key: BiasAdd
[ INFO ]  New subclass: <class 'extensions.front.tf.elementwise_ext.MulExtractor'>
[ INFO ]  Registered a new subclass with key: Mul
[ INFO ]  New subclass: <class 'extensions.front.tf.elementwise_ext.SubExtractor'>
[ INFO ]  Registered a new subclass with key: Sub
[ INFO ]  New subclass: <class 'extensions.front.tf.elementwise_ext.DivExtractor'>
[ INFO ]  Registered a new subclass with key: RealDiv
[ INFO ]  New subclass: <class 'extensions.front.tf.elementwise_ext.SqrtExtractor'>
[ INFO ]  Registered a new subclass with key: Sqrt
[ INFO ]  New subclass: <class 'extensions.front.tf.elementwise_ext.RsqrtExtractor'>
[ INFO ]  Registered a new subclass with key: Rsqrt
[ INFO ]  New subclass: <class 'extensions.front.tf.elementwise_ext.SquareExtractor'>
[ INFO ]  Registered a new subclass with key: Square
[ INFO ]  New subclass: <class 'extensions.front.tf.elementwise_ext.NegExtractor'>
[ INFO ]  Registered a new subclass with key: Neg
[ INFO ]  New subclass: <class 'extensions.front.tf.elementwise_ext.ZerosLike'>
[ INFO ]  Registered a new subclass with key: ZerosLike
[ INFO ]  New subclass: <class 'extensions.front.tf.elementwise_ext.MaximumExtractor'>
[ INFO ]  Registered a new subclass with key: Maximum
[ INFO ]  New subclass: <class 'extensions.front.tf.elementwise_ext.MinimumExtractor'>
[ INFO ]  Registered a new subclass with key: Minimum
[ INFO ]  New subclass: <class 'extensions.front.tf.elementwise_ext.PowExtractor'>
[ INFO ]  Registered a new subclass with key: Pow
[ INFO ]  New subclass: <class 'extensions.front.tf.elementwise_ext.LogicalAndFrontExtractor'>
[ INFO ]  Registered a new subclass with key: LogicalAnd
[ INFO ]  New subclass: <class 'extensions.front.tf.elementwise_ext.LogicalOrFrontExtractor'>
[ INFO ]  Registered a new subclass with key: LogicalOr
[ INFO ]  New subclass: <class 'extensions.front.tf.elementwise_ext.EqualExtractor'>
[ INFO ]  Registered a new subclass with key: Equal
[ INFO ]  New subclass: <class 'extensions.front.tf.elementwise_ext.LessEqualExtractor'>
[ INFO ]  Registered a new subclass with key: LessEqual
[ INFO ]  New subclass: <class 'extensions.front.tf.elementwise_ext.LessExtractor'>
[ INFO ]  Registered a new subclass with key: Less
[ INFO ]  New subclass: <class 'extensions.front.tf.elementwise_ext.GreaterExtractor'>
[ INFO ]  Registered a new subclass with key: Greater
[ INFO ]  New subclass: <class 'extensions.front.tf.elementwise_ext.GreaterEqualExtractor'>
[ INFO ]  Registered a new subclass with key: GreaterEqual
[ INFO ]  New subclass: <class 'extensions.front.tf.elementwise_ext.NotEqualExtractor'>
[ INFO ]  Registered a new subclass with key: NotEqual
[ INFO ]  New subclass: <class 'extensions.front.tf.elu_ext.EluFrontExtractor'>
[ INFO ]  Registered a new subclass with key: Elu
[ INFO ]  New subclass: <class 'extensions.front.tf.erf_ext.ErfFrontExtractor'>
[ INFO ]  Registered a new subclass with key: Erf
[ INFO ]  New subclass: <class 'extensions.front.tf.exp_ext.ExpExtractor'>
[ INFO ]  Registered a new subclass with key: Exp
[ INFO ]  New subclass: <class 'extensions.front.tf.expand_dims_ext.ExpandDimsExtractor'>
[ INFO ]  Registered a new subclass with key: ExpandDims
[ INFO ]  New subclass: <class 'extensions.front.tf.extract_image_patches_ext.ExtractImagePatchesExtractor'>
[ INFO ]  Registered a new subclass with key: ExtractImagePatches
[ INFO ]  New subclass: <class 'extensions.front.tf.fifo_queue_v2_ext.FIFOQueueV2Extractor'>
[ INFO ]  Registered a new subclass with key: FIFOQueueV2
[ INFO ]  New subclass: <class 'extensions.front.tf.fill_ext.FillExtractor'>
[ INFO ]  Registered a new subclass with key: Fill
[ INFO ]  New subclass: <class 'extensions.front.tf.gather_ext.GatherFrontExtractor'>
[ INFO ]  Registered a new subclass with key: Gather
[ INFO ]  New subclass: <class 'extensions.front.tf.gather_ext.ResourceGatherFrontExtractor'>
[ INFO ]  Registered a new subclass with key: ResourceGather
[ INFO ]  New subclass: <class 'extensions.front.tf.gather_ext.GatherV2FrontExtractor'>
[ INFO ]  Registered a new subclass with key: GatherV2
[ INFO ]  New subclass: <class 'extensions.front.tf.leaky_relu_ext.LeakyReLUFrontExtractor'>
[ INFO ]  Registered a new subclass with key: LeakyRelu
[ INFO ]  New subclass: <class 'extensions.front.tf.max_ext.MaxFrontExtractor'>
[ INFO ]  Registered a new subclass with key: Max
[ INFO ]  New subclass: <class 'extensions.front.tf.mean_ext.MeanExtractor'>
[ INFO ]  Registered a new subclass with key: Mean
[ INFO ]  New subclass: <class 'extensions.front.tf.next_iteration_ext.NextIterationExtractor'>
[ INFO ]  Registered a new subclass with key: NextIteration
[ INFO ]  New subclass: <class 'extensions.front.tf.one_hot_ext.OneHotFrontExtractor'>
[ INFO ]  Registered a new subclass with key: OneHot
[ INFO ]  New subclass: <class 'extensions.front.tf.pad_ext.PadFrontExtractor'>
[ INFO ]  Registered a new subclass with key: Pad
[ INFO ]  New subclass: <class 'extensions.front.tf.pad_ext.PadV2FrontExtractor'>
[ INFO ]  Registered a new subclass with key: PadV2
[ INFO ]  New subclass: <class 'extensions.front.tf.pad_ext.MirrorPadFrontExtractor'>
[ INFO ]  Registered a new subclass with key: MirrorPad
[ INFO ]  New subclass: <class 'extensions.front.tf.placeholder_ext.PlaceholderFrontExtractor'>
[ INFO ]  Registered a new subclass with key: Placeholder
[ INFO ]  New subclass: <class 'extensions.front.tf.pooling_ext.AvgPoolFrontExtractor'>
[ INFO ]  Registered a new subclass with key: AvgPool
[ INFO ]  New subclass: <class 'extensions.front.tf.pooling_ext.MaxPoolFrontExtractor'>
[ INFO ]  Registered a new subclass with key: MaxPool
[ INFO ]  New subclass: <class 'extensions.front.tf.pooling_ext.MaxPool3DFrontExtractor'>
[ INFO ]  Registered a new subclass with key: MaxPool3D
[ INFO ]  New subclass: <class 'extensions.front.tf.pooling_ext.AvgPool3DFrontExtractor'>
[ INFO ]  Registered a new subclass with key: AvgPool3D
[ INFO ]  New subclass: <class 'extensions.front.tf.prod_ext.ProdFrontExtractor'>
[ INFO ]  Registered a new subclass with key: Prod
[ INFO ]  New subclass: <class 'extensions.front.tf.rank_ext.RankFrontExtractor'>
[ INFO ]  Registered a new subclass with key: Rank
[ INFO ]  New subclass: <class 'extensions.front.tf.relu6_ext.Relu6FrontExtractor'>
[ INFO ]  Registered a new subclass with key: Relu6
[ INFO ]  New subclass: <class 'extensions.front.tf.relu_ext.ReluFrontExtractor'>
[ INFO ]  Registered a new subclass with key: Relu
[ INFO ]  New subclass: <class 'extensions.front.tf.reshape_ext.ReshapeExtractor'>
[ INFO ]  Registered a new subclass with key: Reshape
[ INFO ]  New subclass: <class 'extensions.front.tf.resize_bilinear.ResizeBilinearFrontExtractor'>
[ INFO ]  Registered a new subclass with key: ResizeBilinear
[ INFO ]  New subclass: <class 'extensions.front.tf.resize_nearest_neighbor.ResizeNearestNeighborFrontExtractor'>
[ INFO ]  Registered a new subclass with key: ResizeNearestNeighbor
[ INFO ]  New subclass: <class 'extensions.front.tf.reverse_sequence.ReverseSequenceFrontExtractor'>
[ INFO ]  Registered a new subclass with key: ReverseSequence
[ INFO ]  New subclass: <class 'extensions.front.tf.reverse_v2.ReverseV2FrontExtractor'>
[ INFO ]  Registered a new subclass with key: ReverseV2
[ INFO ]  New subclass: <class 'extensions.front.tf.select_ext.SelectExtractor'>
[ INFO ]  Registered a new subclass with key: Select
[ INFO ]  New subclass: <class 'extensions.front.tf.shape_ext.ShapeExtractor'>
[ INFO ]  Registered a new subclass with key: Shape
[ INFO ]  New subclass: <class 'extensions.front.tf.sigmoid.SigmoidFrontExtractor'>
[ INFO ]  Registered a new subclass with key: Sigmoid
[ INFO ]  New subclass: <class 'extensions.front.tf.size_ext.SizeFrontExtractor'>
[ INFO ]  Registered a new subclass with key: Size
[ INFO ]  New subclass: <class 'extensions.front.tf.slice_ext.SliceExtractor'>
[ INFO ]  Registered a new subclass with key: Slice
[ INFO ]  New subclass: <class 'extensions.front.tf.softmax_ext.SoftmaxFrontExtractor'>
[ INFO ]  Registered a new subclass with key: Softmax
[ INFO ]  New subclass: <class 'extensions.front.tf.space_to_depth_ext.SpaceToDepthFrontExtractor'>
[ INFO ]  Registered a new subclass with key: SpaceToDepth
[ INFO ]  New subclass: <class 'extensions.front.tf.split_ext.SplitVExtractor'>
[ INFO ]  Registered a new subclass with key: SplitV
[ INFO ]  New subclass: <class 'extensions.front.tf.squeeze_ext.SqueezeExtractor'>
[ INFO ]  Registered a new subclass with key: Squeeze
[ INFO ]  New subclass: <class 'extensions.front.tf.stop_gradient_ext.StopGradientExtractor'>
[ INFO ]  Registered a new subclass with key: StopGradient
[ INFO ]  New subclass: <class 'extensions.front.tf.sum_ext.SumFrontExtractor'>
[ INFO ]  Registered a new subclass with key: Sum
[ INFO ]  New subclass: <class 'extensions.front.tf.tanh.TanhFrontExtractor'>
[ INFO ]  Registered a new subclass with key: Tanh
[ INFO ]  New subclass: <class 'extensions.front.tf.tile_ext.TileExtractor'>
[ INFO ]  Registered a new subclass with key: Tile
[ INFO ]  New subclass: <class 'extensions.front.tf.topk_ext.TopKExtractor'>
[ INFO ]  Registered a new subclass with key: TopK
[ INFO ]  New subclass: <class 'extensions.front.tf.topk_ext.TopKV2Extractor'>
[ INFO ]  Registered a new subclass with key: TopKV2
[ INFO ]  New subclass: <class 'extensions.front.tf.transpose_ext.TransposeFrontExtractorTF'>
[ INFO ]  Registered a new subclass with key: Transpose
[ INFO ]  New subclass: <class 'extensions.front.tf.variable_ext.VariableExtractor'>
[ INFO ]  Registered a new subclass with key: Variable
[ INFO ]  New subclass: <class 'extensions.front.tf.variable_ext.VariableV2Extractor'>
[ INFO ]  Registered a new subclass with key: VariableV2
[ WARNING ]  Skipped <class 'extensions.front.InterpolateNormalizer.InterpolateNormalizer'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.LRNReplacer.LRNReplacer'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.Pack.Pack'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.PowerToEltwises.PowerToEltwises'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.tf.FakeQuantWithMinMaxVars.FakeQuantWithMinMaxVarsToQuantize'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.div.Div'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.eltwise_n.EltwiseNReplacement'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.image_scaler.ImageScaler'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.instance_normalization.InstanceNormalization'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.reciprocal.ReciprocalReplacer'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.sub.Sub'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.squared_difference.SquaredDifference'> registration because it was already registered or it was disabled. 
[ INFO ]  New subclass: <class 'extensions.front.tf.BlockLSTM.BlockLSTM'>
[ INFO ]  Registered a new subclass with key: BlockLSTM
[ INFO ]  New subclass: <class 'extensions.front.tf.CropAndResizeReplacement.CropAndResizeReplacement'>
[ INFO ]  Registered a new subclass with key: CropAndResize
[ INFO ]  New subclass: <class 'extensions.front.tf.LogSoftmax.LogSoftmaxFrontReplacer'>
[ INFO ]  Registered a new subclass with key: LogSoftmax
[ INFO ]  New subclass: <class 'extensions.front.tf.Unpack.Unpack'>
[ INFO ]  Registered a new subclass with key: Unpack
[ INFO ]  New subclass: <class 'extensions.front.tf.SizeReplacer.SizeFrontReplacer'>
[ INFO ]  Registered a new subclass with key: Size
[ INFO ]  New subclass: <class 'extensions.front.tf.assign_elimination.AssignElimination'>
[ INFO ]  Registered a new subclass with key: Assign
[ INFO ]  New subclass: <class 'extensions.front.tf.assign_elimination.AssignSubElimination'>
[ INFO ]  Registered a new subclass with key: AssignSub
[ INFO ]  New subclass: <class 'extensions.front.tf.assign_elimination.AssignAddElimination'>
[ INFO ]  Registered a new subclass with key: AssignAdd
[ INFO ]  New subclass: <class 'extensions.front.tf.assign_elimination.AssertElimination'>
[ INFO ]  Registered a new subclass with key: Assert
[ INFO ]  New subclass: <class 'extensions.front.tf.fake_const_ext.FakeConstToConst'>
[ INFO ]  Registered a new subclass with key: FakeConst
[ WARNING ]  Skipped <class 'mo.front.common.replacement.FrontReplacementSubgraph'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.front.tf.replacement.FrontReplacementFromConfigFileGeneral'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.ExpandDimsToUnsqueeze.ExpandDimsToUnsqueeze'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.FillToBroadcast.FillToBroadcast'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.SqueezeNormalize.SqueezeNormalize'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.TopKNormalize.TopKNormalize'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.binary_quantize_normalization.BinaryFakeQuantizeNormalization'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.create_tensor_nodes.CreateTensorNodes'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.disable_weights_quantize_value_propagation.DisableQuantizeValuePropagation'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.input_cut.InputCut'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.output_cut.OutputCut'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.override_batch.OverrideBatch'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.pass_separator.FrontStart'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.pass_separator.FrontFinish'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.user_data_repack.UserDataRepack'> registration because it was already registered or it was disabled. 
[ INFO ]  New subclass: <class 'extensions.front.tf.GNMT_DynamicSequenceLengths.GNMT_sequence_lengths'>
[ INFO ]  New subclass: <class 'extensions.front.tf.tensorflow_custom_operations_config_update.TensorflowCustomOperationsConfigUpdate'>
[ INFO ]  New subclass: <class 'extensions.front.tf.tensorflow_patterns.TensorflowSubgraphPatterns'>
[ INFO ]  New subclass: <class 'extensions.front.tf.tensorflow_patterns.TensorflowOperationPatterns'>
[ INFO ]  New subclass: <class 'extensions.front.tf.tensorflow_use_custom_operations_config.TensorflowUseCustomOperationsConfig'>
[ INFO ]  New subclass: <class 'extensions.front.tf.variables_values_freezing.VariablesToConstants'>
[ WARNING ]  Skipped <class 'mo.front.common.replacement.FrontReplacementOp'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'mo.front.tf.replacement.FrontReplacementFromConfigFileSubGraph'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.ArgMaxSqueeze.ArgMaxSqueeze'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.TransposeOrderNormalizer.TransposeOrderNormalizer'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.freeze_placeholder_value.FreezePlaceholderValue'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.no_op_eraser.NoOpEraser'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.reduce_axis_normalizer.ReduceAxisNormalizer'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.reshape_dim_normalizer.ReshapeDimNormalizer'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.restore_ports.RestorePorts'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.softmax.SoftmaxFromKeras'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.front.standalone_const_eraser.StandaloneConstEraser'> registration because it was already registered or it was disabled. 
[ INFO ]  New subclass: <class 'extensions.front.tf.CTCGreedyDecoder.CTCGreedyDecoderReplacement'>
[ INFO ]  New subclass: <class 'extensions.front.tf.nearest_neighbor_upsampling.NearestNeighborUpsampling'>
[ INFO ]  New subclass: <class 'extensions.front.tf.FlattenToReshape.FlattenToReshapeableReshape'>
[ INFO ]  New subclass: <class 'extensions.front.tf.basic_lstm_cell.BasicLSTMCell'>
[ INFO ]  New subclass: <class 'extensions.front.tf.concat.Concat'>
[ INFO ]  New subclass: <class 'extensions.front.tf.fifo_replacer.FIFOQueue'>
[ INFO ]  New subclass: <class 'extensions.front.tf.fifo_replacer.QueueDequeueManyV2'>
[ INFO ]  New subclass: <class 'extensions.front.tf.mvn.MVN'>
[ INFO ]  New subclass: <class 'extensions.front.tf.mvn_unrolled.MVNUnrolled'>
[ INFO ]  New subclass: <class 'extensions.front.tf.prelu.PReLU'>
[ INFO ]  New subclass: <class 'extensions.front.tf.prelu.PReLUWithAbs'>
[ WARNING ]  Skipped <class 'mo.front.tf.replacement.FrontReplacementFromConfigFileOp'> registration because it was already registered or it was disabled. 
[ INFO ]  New subclass: <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIPreprocessorReplacement'>
[ INFO ]  New subclass: <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIDetectionOutputReplacement'>
[ INFO ]  New subclass: <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIMaskRCNNROIPoolingSecondReplacement'>
[ INFO ]  New subclass: <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIProposalReplacement'>
[ INFO ]  New subclass: <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPISSDPostprocessorReplacement'>
[ INFO ]  New subclass: <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIPSROIPoolingReplacement'>
[ INFO ]  New subclass: <class 'extensions.front.tf.RetinaNetFilteredDetectionsReplacement.RetinaNetFilteredDetectionsReplacement'>
[ INFO ]  New subclass: <class 'extensions.front.tf.SSDToolboxDetectionOutput.SSDToolboxDetectionOutputReplacement'>
[ INFO ]  New subclass: <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIMaskRCNNSigmoidReplacement'>
[ INFO ]  New subclass: <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIOutputReplacement'>
[ INFO ]  New subclass: <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIConstValueOverride'>
[ INFO ]  New subclass: <class 'extensions.front.tf.YOLO.YoloRegionAddon'>
[ INFO ]  New subclass: <class 'extensions.front.tf.YOLO.YoloV3RegionAddon'>
[ WARNING ]  Skipped <class 'extensions.middle.ConcatOptimization.ConcatOptimization'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.middle.EltwiseChecker.EltwiseChecker'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.middle.EltwiseInputReshape.Eltwise1DInputReshape'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.middle.EltwiseInputReshape.EltwiseInputReshape'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.middle.EltwiseInputNormalization.EltwiseInputNormalize'> registration because it was already registered or it was disabled. 
[ INFO ]  New subclass: <class 'extensions.middle.PartialInfer.PartialInfer'>
[ INFO ]  New subclass: <class 'extensions.middle.TensorIteratorCondition.LoopConditionMatcher'>
[ INFO ]  New subclass: <class 'extensions.middle.TensorIteratorCondition.SimpleConditionMatcher'>
[ INFO ]  New subclass: <class 'extensions.middle.TensorIteratorCondition.DynamicDecoderConditionMatcher'>
[ INFO ]  New subclass: <class 'extensions.middle.MulFakeQuantizeFuse.MulFakeQuantizeFuse'>
[ INFO ]  New subclass: <class 'extensions.middle.AddFakeQuantizeFuse.AddFakeQuantizeFuse'>
[ INFO ]  New subclass: <class 'extensions.middle.AddIsCyclicAttribute.AddIsCyclicAttribute'>
[ INFO ]  New subclass: <class 'extensions.middle.AddMeanScaleValues.AddMeanScaleValues'>
[ INFO ]  New subclass: <class 'extensions.middle.AnchorToPriorBox.AnchorToPriorBoxes'>
[ INFO ]  New subclass: <class 'extensions.middle.ArgMaxToTopK.ArgMaxToTopK'>
[ INFO ]  Registered a new subclass with key: ArgMax
[ INFO ]  New subclass: <class 'extensions.middle.CheckForCycle.CheckForCycle'>
[ INFO ]  New subclass: <class 'extensions.middle.DeleteNotExecutable.DeleteNotExecutable'>
[ INFO ]  New subclass: <class 'extensions.middle.BinarizeWeightsM1P1.BinarizeWeightsM1P1'>
[ INFO ]  New subclass: <class 'extensions.middle.BlockLSTMtoLSTMSequence.BlockLSTMtoLSTMSequence'>
[ INFO ]  New subclass: <class 'extensions.middle.RemoveIdentity.RemoveIdentity'>
[ INFO ]  New subclass: <class 'extensions.middle.RemoveIdentity.RemoveDropout'>
[ INFO ]  New subclass: <class 'extensions.middle.RemoveIdentity.RemoveNodesWithZeroPhase'>
[ INFO ]  New subclass: <class 'extensions.middle.Cast.CastToFloatMark'>
[ INFO ]  New subclass: <class 'extensions.middle.ChangePlaceholderTypes.ChangePlaceholderTypes'>
[ INFO ]  New subclass: <class 'extensions.middle.ConstSwitchResolver.ConstSwitchEraser'>
[ INFO ]  New subclass: <class 'extensions.middle.ConvToBinaryConv.ConvToBinaryConv'>
[ INFO ]  New subclass: <class 'extensions.middle.SliceConverter.ConvertSlice'>
[ INFO ]  Registered a new subclass with key: Slice
[ INFO ]  New subclass: <class 'extensions.middle.ConvertGroupedStridedSlice.ConvertGroupedStridedSlice'>
[ INFO ]  New subclass: <class 'extensions.middle.ConvertLayoutDependentOperations.ConvertLayoutDependentOperations'>
[ INFO ]  New subclass: <class 'extensions.middle.ConvertMultiInputConv.ConvertMultiInputConv'>
[ INFO ]  New subclass: <class 'extensions.middle.CustomSubgraphCall.CustomSubgraphCall'>
[ INFO ]  New subclass: <class 'extensions.middle.DecomposeBias.DecomposeBias'>
[ INFO ]  New subclass: <class 'extensions.middle.DecomposeBidirectionalRNNSequence.DecomposeBidirectionalRNNSequence'>
[ INFO ]  New subclass: <class 'extensions.middle.DeleteControlFlowEdges.DeleteControlFlowEdges'>
[ INFO ]  New subclass: <class 'extensions.middle.DepthToSpace.DepthToSpace'>
[ INFO ]  New subclass: <class 'extensions.middle.DilatedConvolution.DilatedConvolutionConverter'>
[ INFO ]  New subclass: <class 'extensions.middle.TensorIteratorMerge.TensorIteratorMerge'>
[ INFO ]  New subclass: <class 'extensions.middle.FakeSplitOutputs.AddFakeOutputsToSplit'>
[ INFO ]  New subclass: <class 'extensions.middle.FlattenONNXToReshape.FlattenONNXToReshape'>
[ INFO ]  New subclass: <class 'extensions.middle.FusedBatchNormNonConstant.FusedBatchNormNonConstant'>
[ INFO ]  New subclass: <class 'extensions.middle.FusedBatchNormTrainingCatch.FusedBatchNormTrainingCatch'>
[ INFO ]  New subclass: <class 'extensions.middle.GRURNNSequenceToTensorIterator.GRUAndRNNToTensorIterator'>
[ INFO ]  New subclass: <class 'extensions.middle.GatherNdNormalizer.GatherNdNormalize'>
[ INFO ]  New subclass: <class 'extensions.middle.GemmResolver.GemmResolver'>
[ INFO ]  New subclass: <class 'extensions.middle.GemmToFullyConnected.GemmToFullyConnected'>
[ INFO ]  New subclass: <class 'extensions.middle.GroupNorm.GroupNormToMVN'>
[ INFO ]  Registered a new subclass with key: GroupNorm
[ INFO ]  New subclass: <class 'extensions.middle.InputCut.MiddleInputCut'>
[ INFO ]  New subclass: <class 'extensions.middle.L2NormToNorm.L2NormToNorm'>
[ INFO ]  New subclass: <class 'extensions.middle.RNNSequenceNormalizeToIE.RNNSequenceNormalize'>
[ INFO ]  New subclass: <class 'extensions.middle.LSTMRNNSequenceToTensorIterator.LSTMToTensorIterator'>
[ INFO ]  New subclass: <class 'extensions.middle.MXNetRNNSequenceNormalize.MXNetRNNSequenceNormalize'>
[ INFO ]  New subclass: <class 'extensions.middle.MXNetSplitMultiLayers.MXNetSplitLayersToRNNSequence'>
[ INFO ]  New subclass: <class 'extensions.middle.MinimumMiddleReplacer.MinimumMiddleReplacer'>
[ INFO ]  New subclass: <class 'extensions.middle.NasNet.NasNet'>
[ INFO ]  New subclass: <class 'extensions.middle.NormalizeFullyConnected.NormalizeFullyConnected'>
[ INFO ]  New subclass: <class 'extensions.middle.NormalizePad.NormalizePad'>
[ INFO ]  New subclass: <class 'extensions.middle.ONNXRNNSequenceNormalize.ONNXRNNSequenceNormalize'>
[ INFO ]  New subclass: <class 'extensions.middle.ReluQuantizeFuse.ReluFakeQuantizeMark'>
[ INFO ]  New subclass: <class 'extensions.middle.ReluQuantizeFuse.ClampQuantizeMark'>
[ INFO ]  New subclass: <class 'extensions.middle.ReluQuantizeFuse.ReluQuantizeFuse'>
[ INFO ]  New subclass: <class 'extensions.middle.RemoveDuplicationMemory.RemoveMemoryDuplicationPattern'>
[ INFO ]  New subclass: <class 'extensions.middle.RemoveRedundantReshapeAfterCropAndResize.RemoveRedundantReshapeAfterCropAndResize'>
[ INFO ]  New subclass: <class 'extensions.middle.RemoveUselessCrops.RemoveUselessCropsPattern'>
[ INFO ]  New subclass: <class 'extensions.middle.ReplaceMemoryOffsetWithSplice.ReplaceMemoryOffsetNodePattern'>
[ INFO ]  New subclass: <class 'extensions.middle.ReplaceMemoryOffsetWithSplice.ReplaceMemoryOffsetWithMemoryNodePattern'>
[ INFO ]  New subclass: <class 'extensions.middle.ReplaceSpliceNodePattern.ReplaceSpliceNodePattern'>
[ INFO ]  New subclass: <class 'extensions.middle.ReverseTransposeNormalization.ReverseTransposeNormalization'>
[ INFO ]  New subclass: <class 'extensions.middle.ReverseV2ToReverseSequence.ReverseToReverseSequence'>
[ INFO ]  New subclass: <class 'extensions.middle.ScaleInput.ScaleInput'>
[ INFO ]  New subclass: <class 'extensions.middle.SharedWeightsDuplication.SharedWeightsDuplication'>
[ INFO ]  New subclass: <class 'extensions.middle.ShuffleChannel.ShuffleChannel'>
[ INFO ]  New subclass: <class 'extensions.middle.SwapAxesMiddleReplacer.SwapAxisMiddleReplacer'>
[ INFO ]  New subclass: <class 'extensions.middle.TensorIteratorInput.SmartInputMatcher'>
[ INFO ]  New subclass: <class 'extensions.middle.TensorIteratorInput.SimpleInputMatcher'>
[ INFO ]  New subclass: <class 'extensions.middle.TensorIteratorInput.BackEdgeSimpleInputMatcher'>
[ INFO ]  New subclass: <class 'extensions.middle.SwitchMergeOptimization.SwitchMergeMiddleReplacer'>
[ INFO ]  New subclass: <class 'extensions.middle.TF_lstm_cell_to_generic.TensorFlowLSTMtoGeneric'>
[ INFO ]  New subclass: <class 'extensions.middle.TensorIteratorBackEdge.BackEdgesMatching'>
[ INFO ]  New subclass: <class 'extensions.middle.TensorIteratorConditionChecker.ConditionChecks'>
[ INFO ]  New subclass: <class 'extensions.middle.TensorIteratorLSTMToLSTMSequence.TensorIteratorLSTM'>
[ INFO ]  New subclass: <class 'extensions.middle.TensorIteratorOutput.SmartOutputMatcher'>
[ INFO ]  New subclass: <class 'extensions.middle.TensorIteratorOutput.SimpleOutputMatcher'>
[ INFO ]  New subclass: <class 'extensions.middle.TileReplacer.TileMultipleAxisReplacer'>
[ INFO ]  New subclass: <class 'extensions.middle.UpsampleToResample.UpsampleToResample'>
[ INFO ]  New subclass: <class 'extensions.middle.UselessMerge.UselessMergeEraser'>
[ INFO ]  New subclass: <class 'extensions.middle.UselessSplitEraser.UselessSplitEraser'>
[ INFO ]  New subclass: <class 'extensions.middle.UselessStridedSlice.UselessStridedSliceEraser'>
[ INFO ]  New subclass: <class 'extensions.middle.pass_separator.PreMiddleStart'>
[ INFO ]  New subclass: <class 'extensions.middle.pass_separator.MiddleStart'>
[ INFO ]  New subclass: <class 'extensions.middle.pass_separator.MiddleFinish'>
[ INFO ]  New subclass: <class 'extensions.middle.permute_tensor_iterator.TransposeTensorIteratorLSTM'>
[ INFO ]  New subclass: <class 'extensions.middle.quantize_fuses.MarkNodesToFuseUpToFakeQuantize'>
[ INFO ]  New subclass: <class 'extensions.middle.quantize_fuses.FakeQuantizeFuse'>
[ INFO ]  New subclass: <class 'extensions.middle.reverse_tensor_iterator.ReverseTensorIteratorLSTM'>
[ INFO ]  New subclass: <class 'extensions.middle.space_to_depth.SpaceToDepth'>
[ INFO ]  New subclass: <class 'extensions.middle.ssd_anchors_to_const.SsdAnchorsMiddleReplacer'>
[ INFO ]  New subclass: <class 'extensions.middle.wights_permute_normalizer.WeightsPermuteNormalizer'>
[ WARNING ]  Skipped <class 'extensions.back.CreateConstNodes.CreateConstNodesReplacement'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.back.FuseReshapesSequence.FuseReshapesSequence'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.back.InsertLayoutPropagationTransposes.InsertLayoutPropagationTranspose'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.back.RemoveRedundantReshapes.RemoveRedundantReshapes'> registration because it was already registered or it was disabled. 
[ WARNING ]  Skipped <class 'extensions.middle.LayoutChangeForConstantShapePaths.LayoutChangeForConstantShapePaths'> registration because it was already registered or it was disabled. 
[ INFO ]  New subclass: <class 'extensions.back.ActivationsNormalizer.ActivationsNormalizer'>
[ INFO ]  New subclass: <class 'extensions.back.AvgPool.AvgPool'>
[ INFO ]  New subclass: <class 'extensions.back.ConvolutionNormalizer.ConvolutionNormalizer'>
[ INFO ]  New subclass: <class 'extensions.back.EltwiseBroadcast.EltwiseBroadcast'>
[ INFO ]  New subclass: <class 'extensions.back.ForceStrictPrecision.ForceStrictPrecision'>
[ INFO ]  New subclass: <class 'extensions.back.FuseTransposesSequence.FuseTransposesSequence'>
[ INFO ]  New subclass: <class 'extensions.back.OptimizeTransposeReshapeSequence.OptimizeTransposeReshapeSequence'>
[ INFO ]  New subclass: <class 'extensions.back.ReshapeMutation.ReshapeMutation'>
[ INFO ]  New subclass: <class 'extensions.back.ReshapeMutation.FlattenMutation'>
[ INFO ]  New subclass: <class 'extensions.back.ReshapeMutation.DisableReshapeMutationInTensorIterator'>
[ INFO ]  New subclass: <class 'extensions.back.ConvolutionReshaper.ConvolutionReshaper'>
[ INFO ]  New subclass: <class 'extensions.back.CropToStridedSlice.CropToStridedSlice'>
[ INFO ]  New subclass: <class 'extensions.back.DumpFakeQuantStat.DumpFakeQuantStat'>
[ INFO ]  New subclass: <class 'extensions.back.ElementwiseOpsToEltwiseOps.SimpleEltwiseToEltwiseOp'>
[ INFO ]  New subclass: <class 'extensions.back.ElementwiseOpsToEltwiseOps.DivideToEltwises'>
[ INFO ]  New subclass: <class 'extensions.back.ElementwiseOpsToEltwiseOps.SubtractToEltwises'>
[ INFO ]  New subclass: <class 'extensions.back.ElementwiseOpsToEltwiseOps.EltwisesWithScalarInputToPower'>
[ INFO ]  New subclass: <class 'extensions.back.ElementwiseOpsToEltwiseOps.MulAddPowerMerge'>
[ INFO ]  New subclass: <class 'extensions.back.ElementwiseOpsToEltwiseOps.MulPowPowerMerge'>
[ INFO ]  New subclass: <class 'extensions.back.ElementwiseOpsToEltwiseOps.AddPowPowerMerge'>
[ INFO ]  New subclass: <class 'extensions.back.ElementwiseOpsToEltwiseOps.MulAddPowPowerMerge'>
[ INFO ]  New subclass: <class 'extensions.back.EnableConstantStridedSlice.EnableConstantStridedSlice'>
[ INFO ]  New subclass: <class 'extensions.back.GroupedConvWeightsNormalize.GroupedConvWeightsNormalize'>
[ INFO ]  New subclass: <class 'extensions.back.InterpolateToInterpOrResample.InterpolateToInterpOrResample'>
[ INFO ]  New subclass: <class 'extensions.back.LRNToNorm.LRNToNorm'>
[ INFO ]  New subclass: <class 'extensions.back.LeakyReluToReluWithNegativeSlope.LeakyReluToReluWithNegativeSlope'>
[ INFO ]  New subclass: <class 'extensions.back.MatMulNormalizer.MatMulNormalizer'>
[ INFO ]  New subclass: <class 'extensions.back.MaxPool.MaxPool'>
[ INFO ]  New subclass: <class 'extensions.back.NormalizeToNormalizeL2.NormalizeToNormalizeL2'>
[ INFO ]  New subclass: <class 'extensions.back.PackBinaryWeights.PackBinaryWeights'>
[ INFO ]  New subclass: <class 'extensions.back.PadNormalizer.PadNormalize'>
[ INFO ]  New subclass: <class 'extensions.back.ParameterToPlaceholder.ParameterToInput'>
[ INFO ]  New subclass: <class 'extensions.back.RNNSequenceTypeRename.RNNSequence'>
[ INFO ]  New subclass: <class 'extensions.back.ReduceToPooling.ReduceReplacer'>
[ INFO ]  New subclass: <class 'extensions.back.ReduceTransposeDimensions.ReduceTransposeDimensions'>
[ INFO ]  New subclass: <class 'extensions.back.RepackFCWeightsNHWCToNCHW.RepackFCWeightsNHWCToNCHW'>
[ INFO ]  New subclass: <class 'extensions.back.ResultNormalizer.ResultNormalizer'>
[ INFO ]  New subclass: <class 'extensions.back.ScalarConstNormalize.ScalarNormalize'>
[ INFO ]  New subclass: <class 'extensions.back.ScalarConstNormalize.ScalarNormalizeForSpecificOps'>
[ INFO ]  New subclass: <class 'extensions.back.SelectBroadcast.SelectBroadcast'>
[ INFO ]  New subclass: <class 'extensions.back.SelectInputNormalizer.SelectInputNormalizer'>
[ INFO ]  New subclass: <class 'extensions.back.ShapeOfToShape.ShapeOfToShape'>
[ INFO ]  New subclass: <class 'extensions.back.TransposeToPermute.TransposeToPermute'>
[ INFO ]  New subclass: <class 'extensions.back.ShuffleChannelPatternOptimization.ShuffleChannelPatternOptimization'>
[ INFO ]  New subclass: <class 'extensions.back.ShufflenetReLUReorder.ShufflenetReLUReorder'>
[ INFO ]  New subclass: <class 'extensions.back.TileReshaper.TileReshaper'>
[ INFO ]  New subclass: <class 'extensions.back.TopKNormalizer.TopKNormalizer'>
[ INFO ]  New subclass: <class 'extensions.back.TransposeFCWeights.TransposeFCWeights'>
[ INFO ]  New subclass: <class 'extensions.back.blob_normalizer.BlobNormalizer'>
[ INFO ]  New subclass: <class 'extensions.back.disable_unsupported_ND_operations.DisableUnsupportedNDOperations'>
[ INFO ]  New subclass: <class 'extensions.back.insert_compatibility_l2normalization.CompatibilityL2NormalizationPattern'>
[ INFO ]  New subclass: <class 'extensions.back.kaldi_remove_memory_output.KaldiRemoveMemoryOutputBackReplacementPattern'>
[ INFO ]  New subclass: <class 'extensions.back.priorbox_mutation.PriorboxMutation'>
[ INFO ]  New subclass: <class 'extensions.back.remove_last_softmax_pattern.RemoveLastSoftMaxPattern'>
[ 2019-08-16 11:41:45,581 ] [ DEBUG ] [ class_registration:48 ]  All replacers has unique idxs.
[ 2019-08-16 11:41:45,637 ] [ DEBUG ] [ tf:93 ]  Number of nodes in graph_def: 755
[ 2019-08-16 11:41:45,637 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry StridedSlice to extractors with custom extractor class <class 'mo.front.tf.extractors.strided_slice.StridedSliceFrontExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:70 ]  Overridden extractor entry BatchMatMul by custom extractor class <class 'extensions.front.tf.BatchMatMul_ext.BatchMatMulExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry BlockLSTM to extractors with custom extractor class <class 'extensions.front.tf.BlockLSTM_ext.BlockLSTMExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry CTCGreedyDecoder to extractors with custom extractor class <class 'extensions.front.tf.CTCGreedyDecoder_ext.CTCCGreedyDecoderFrontExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Cast to extractors with custom extractor class <class 'extensions.front.tf.Cast_ext.CastFrontExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry FakeQuantWithMinMaxVars to extractors with custom extractor class <class 'extensions.front.tf.FakeQuantWithMinMaxVars_ext.FakeQuantWithMinMaxVarsExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry FloorMod to extractors with custom extractor class <class 'extensions.front.tf.FloorMod_ext.EltwiseFrontExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry GatherTree to extractors with custom extractor class <class 'extensions.front.tf.GatherTree_ext.GatherTreeFrontExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry LoopCond to extractors with custom extractor class <class 'extensions.front.tf.LoopCond_ext.LoopCondFrontExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry LogicalNot to extractors with custom extractor class <class 'extensions.front.tf.Not_ext.LogicalNotFrontExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry TensorArrayV3 to extractors with custom extractor class <class 'extensions.front.tf.TensorArrayExtractors.TensorArrayExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry TensorArrayGatherV3 to extractors with custom extractor class <class 'extensions.front.tf.TensorArrayGatherV3.TensorArrayGatherV3Exteractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry All to extractors with custom extractor class <class 'extensions.front.tf.all_ext.AllFrontExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry ArgMax to extractors with custom extractor class <class 'extensions.front.tf.argmax_ext.ArgMaxFrontExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Concat to extractors with custom extractor class <class 'extensions.front.tf.concat_ext.ConcatFrontExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Conv2D to extractors with custom extractor class <class 'extensions.front.tf.conv_ext.Conv2DFrontExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry DepthwiseConv2dNative to extractors with custom extractor class <class 'extensions.front.tf.conv_ext.DepthwiseConv2dNativeFrontExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Conv3D to extractors with custom extractor class <class 'extensions.front.tf.conv_ext.Conv3DFrontExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry CropAndResize to extractors with custom extractor class <class 'extensions.front.tf.crop_and_resize_ext.CropAndResizeFrontExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Conv2DBackpropInput to extractors with custom extractor class <class 'extensions.front.tf.deconv_ext.Conv2DBackpropInputFrontExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Conv3DBackpropInputV2 to extractors with custom extractor class <class 'extensions.front.tf.deconv_ext.Conv3DBackpropInputV2InputFrontExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry DepthToSpace to extractors with custom extractor class <class 'extensions.front.tf.depth_to_space.DepthToSpaceFrontExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Add to extractors with custom extractor class <class 'extensions.front.tf.elementwise_ext.AddExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry AddN to extractors with custom extractor class <class 'extensions.front.tf.elementwise_ext.AddNExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry BiasAdd to extractors with custom extractor class <class 'extensions.front.tf.elementwise_ext.BiasAddExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Mul to extractors with custom extractor class <class 'extensions.front.tf.elementwise_ext.MulExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Sub to extractors with custom extractor class <class 'extensions.front.tf.elementwise_ext.SubExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry RealDiv to extractors with custom extractor class <class 'extensions.front.tf.elementwise_ext.DivExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Sqrt to extractors with custom extractor class <class 'extensions.front.tf.elementwise_ext.SqrtExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Rsqrt to extractors with custom extractor class <class 'extensions.front.tf.elementwise_ext.RsqrtExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Square to extractors with custom extractor class <class 'extensions.front.tf.elementwise_ext.SquareExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Neg to extractors with custom extractor class <class 'extensions.front.tf.elementwise_ext.NegExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry ZerosLike to extractors with custom extractor class <class 'extensions.front.tf.elementwise_ext.ZerosLike'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Maximum to extractors with custom extractor class <class 'extensions.front.tf.elementwise_ext.MaximumExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Minimum to extractors with custom extractor class <class 'extensions.front.tf.elementwise_ext.MinimumExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Pow to extractors with custom extractor class <class 'extensions.front.tf.elementwise_ext.PowExtractor'>.
[ 2019-08-16 11:41:45,638 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry LogicalAnd to extractors with custom extractor class <class 'extensions.front.tf.elementwise_ext.LogicalAndFrontExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry LogicalOr to extractors with custom extractor class <class 'extensions.front.tf.elementwise_ext.LogicalOrFrontExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Equal to extractors with custom extractor class <class 'extensions.front.tf.elementwise_ext.EqualExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry LessEqual to extractors with custom extractor class <class 'extensions.front.tf.elementwise_ext.LessEqualExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Less to extractors with custom extractor class <class 'extensions.front.tf.elementwise_ext.LessExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Greater to extractors with custom extractor class <class 'extensions.front.tf.elementwise_ext.GreaterExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry GreaterEqual to extractors with custom extractor class <class 'extensions.front.tf.elementwise_ext.GreaterEqualExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry NotEqual to extractors with custom extractor class <class 'extensions.front.tf.elementwise_ext.NotEqualExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Elu to extractors with custom extractor class <class 'extensions.front.tf.elu_ext.EluFrontExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Erf to extractors with custom extractor class <class 'extensions.front.tf.erf_ext.ErfFrontExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Exp to extractors with custom extractor class <class 'extensions.front.tf.exp_ext.ExpExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry ExpandDims to extractors with custom extractor class <class 'extensions.front.tf.expand_dims_ext.ExpandDimsExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry ExtractImagePatches to extractors with custom extractor class <class 'extensions.front.tf.extract_image_patches_ext.ExtractImagePatchesExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry FIFOQueueV2 to extractors with custom extractor class <class 'extensions.front.tf.fifo_queue_v2_ext.FIFOQueueV2Extractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Fill to extractors with custom extractor class <class 'extensions.front.tf.fill_ext.FillExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Gather to extractors with custom extractor class <class 'extensions.front.tf.gather_ext.GatherFrontExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry ResourceGather to extractors with custom extractor class <class 'extensions.front.tf.gather_ext.ResourceGatherFrontExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry GatherV2 to extractors with custom extractor class <class 'extensions.front.tf.gather_ext.GatherV2FrontExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry LeakyRelu to extractors with custom extractor class <class 'extensions.front.tf.leaky_relu_ext.LeakyReLUFrontExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Max to extractors with custom extractor class <class 'extensions.front.tf.max_ext.MaxFrontExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Mean to extractors with custom extractor class <class 'extensions.front.tf.mean_ext.MeanExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry NextIteration to extractors with custom extractor class <class 'extensions.front.tf.next_iteration_ext.NextIterationExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry OneHot to extractors with custom extractor class <class 'extensions.front.tf.one_hot_ext.OneHotFrontExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Pad to extractors with custom extractor class <class 'extensions.front.tf.pad_ext.PadFrontExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry PadV2 to extractors with custom extractor class <class 'extensions.front.tf.pad_ext.PadV2FrontExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry MirrorPad to extractors with custom extractor class <class 'extensions.front.tf.pad_ext.MirrorPadFrontExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Placeholder to extractors with custom extractor class <class 'extensions.front.tf.placeholder_ext.PlaceholderFrontExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry AvgPool to extractors with custom extractor class <class 'extensions.front.tf.pooling_ext.AvgPoolFrontExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry MaxPool to extractors with custom extractor class <class 'extensions.front.tf.pooling_ext.MaxPoolFrontExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry MaxPool3D to extractors with custom extractor class <class 'extensions.front.tf.pooling_ext.MaxPool3DFrontExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry AvgPool3D to extractors with custom extractor class <class 'extensions.front.tf.pooling_ext.AvgPool3DFrontExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Prod to extractors with custom extractor class <class 'extensions.front.tf.prod_ext.ProdFrontExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Rank to extractors with custom extractor class <class 'extensions.front.tf.rank_ext.RankFrontExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Relu6 to extractors with custom extractor class <class 'extensions.front.tf.relu6_ext.Relu6FrontExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Relu to extractors with custom extractor class <class 'extensions.front.tf.relu_ext.ReluFrontExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Reshape to extractors with custom extractor class <class 'extensions.front.tf.reshape_ext.ReshapeExtractor'>.
[ 2019-08-16 11:41:45,639 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry ResizeBilinear to extractors with custom extractor class <class 'extensions.front.tf.resize_bilinear.ResizeBilinearFrontExtractor'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry ResizeNearestNeighbor to extractors with custom extractor class <class 'extensions.front.tf.resize_nearest_neighbor.ResizeNearestNeighborFrontExtractor'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry ReverseSequence to extractors with custom extractor class <class 'extensions.front.tf.reverse_sequence.ReverseSequenceFrontExtractor'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry ReverseV2 to extractors with custom extractor class <class 'extensions.front.tf.reverse_v2.ReverseV2FrontExtractor'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Select to extractors with custom extractor class <class 'extensions.front.tf.select_ext.SelectExtractor'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Shape to extractors with custom extractor class <class 'extensions.front.tf.shape_ext.ShapeExtractor'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Sigmoid to extractors with custom extractor class <class 'extensions.front.tf.sigmoid.SigmoidFrontExtractor'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Size to extractors with custom extractor class <class 'extensions.front.tf.size_ext.SizeFrontExtractor'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Slice to extractors with custom extractor class <class 'extensions.front.tf.slice_ext.SliceExtractor'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Softmax to extractors with custom extractor class <class 'extensions.front.tf.softmax_ext.SoftmaxFrontExtractor'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry SpaceToDepth to extractors with custom extractor class <class 'extensions.front.tf.space_to_depth_ext.SpaceToDepthFrontExtractor'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry SplitV to extractors with custom extractor class <class 'extensions.front.tf.split_ext.SplitVExtractor'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Squeeze to extractors with custom extractor class <class 'extensions.front.tf.squeeze_ext.SqueezeExtractor'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry StopGradient to extractors with custom extractor class <class 'extensions.front.tf.stop_gradient_ext.StopGradientExtractor'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Sum to extractors with custom extractor class <class 'extensions.front.tf.sum_ext.SumFrontExtractor'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Tanh to extractors with custom extractor class <class 'extensions.front.tf.tanh.TanhFrontExtractor'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Tile to extractors with custom extractor class <class 'extensions.front.tf.tile_ext.TileExtractor'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry TopK to extractors with custom extractor class <class 'extensions.front.tf.topk_ext.TopKExtractor'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry TopKV2 to extractors with custom extractor class <class 'extensions.front.tf.topk_ext.TopKV2Extractor'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Transpose to extractors with custom extractor class <class 'extensions.front.tf.transpose_ext.TransposeFrontExtractorTF'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry Variable to extractors with custom extractor class <class 'extensions.front.tf.variable_ext.VariableExtractor'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:74 ]  Added a new entry VariableV2 to extractors with custom extractor class <class 'extensions.front.tf.variable_ext.VariableV2Extractor'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry ScaleShift to extractors with custom op class <class 'mo.ops.scale_shift.ScaleShiftOp'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Clamp to extractors with custom op class <class 'mo.ops.clamp.Clamp'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Convolution to extractors with custom op class <class 'mo.ops.convolution.Convolution'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Crop to extractors with custom op class <class 'mo.ops.crop.Crop'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Activation to extractors with custom op class <class 'mo.ops.activation.Activation'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Broadcast to extractors with custom op class <class 'mo.ops.broadcast.Broadcast'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Deconvolution to extractors with custom op class <class 'mo.ops.deconvolution.Deconvolution'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry DeformableConvolution to extractors with custom op class <class 'mo.ops.deformable_convolution.DeformableConvolution'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Eltwise to extractors with custom op class <class 'mo.ops.eltwise.Eltwise'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry EltwiseN to extractors with custom op class <class 'mo.ops.eltwise_n.EltwiseN'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry EltwiseNin1 to extractors with custom op class <class 'mo.ops.eltwise_ninputs_in_1.EltwiseNin1'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Flatten to extractors with custom op class <class 'mo.ops.flatten.Flatten'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry FlattenONNX to extractors with custom op class <class 'mo.ops.flatten_onnx.FlattenONNX'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry GroupNorm to extractors with custom op class <class 'mo.ops.group_norm.GroupNorm'>.
[ 2019-08-16 11:41:45,640 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry FullyConnected to extractors with custom op class <class 'mo.ops.inner_product.InnerProduct'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry LstmNonLinearity to extractors with custom op class <class 'mo.ops.lstmnonlinearity.LstmNonLinearity'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Memory to extractors with custom op class <class 'mo.ops.memory.Memory'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry MemoryOffset to extractors with custom op class <class 'mo.ops.memoryoffset.MemoryOffset'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Permute to extractors with custom op class <class 'mo.ops.permute.Permute'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Pooling to extractors with custom op class <class 'mo.ops.pooling.Pooling'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Power to extractors with custom op class <class 'mo.ops.power.Power'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Result to extractors with custom op class <class 'mo.ops.result.Result'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry ROIPooling to extractors with custom op class <class 'mo.ops.roipooling.ROIPooling'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry ShapeOf to extractors with custom op class <class 'mo.ops.shape.Shape'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Unsqueeze to extractors with custom op class <class 'mo.ops.unsqueeze.Unsqueeze'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry DetectionOutput to extractors with custom op class <class 'extensions.ops.DetectionOutput.DetectionOutput'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Enter to extractors with custom op class <class 'extensions.ops.Enter.Enter'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Exit to extractors with custom op class <class 'extensions.ops.Exit.Exit'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry RNN to extractors with custom op class <class 'extensions.ops.RNN.RNN'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry GRU to extractors with custom op class <class 'extensions.ops.GRU.GRU'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry GRUCell to extractors with custom op class <class 'extensions.ops.GRUCell.GRUCell'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry GatherNd to extractors with custom op class <class 'extensions.ops.GatherNd.GatherNd'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry LSTM to extractors with custom op class <class 'extensions.ops.LSTM.LSTM'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Log to extractors with custom op class <class 'extensions.ops.Log.LogOp'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry RNNCell to extractors with custom op class <class 'extensions.ops.RNNCell.RNNCell'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Reverse to extractors with custom op class <class 'extensions.ops.Reverse.Reverse'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry SquaredDifference to extractors with custom op class <class 'extensions.ops.SquaredDifference.SquaredDifference'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry TensorArrayReadV3 to extractors with custom op class <class 'extensions.ops.TensorArrayRead.TensorArrayReader'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry TensorArrayScatterV3 to extractors with custom op class <class 'extensions.ops.TensorArrayScatter.TensorArrayScatter'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry TensorArraySizeV3 to extractors with custom op class <class 'extensions.ops.TensorArraySize.TensorArraySize'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry TensorArrayWriteV3 to extractors with custom op class <class 'extensions.ops.TensorArrayWrite.TensorArrayWriter'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry TensorIteratorInput to extractors with custom op class <class 'extensions.ops.TensorIterator_ops.TensorIteratorInput'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry TensorIteratorOutput to extractors with custom op class <class 'extensions.ops.TensorIterator_ops.TensorIteratorOutput'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry TensorIteratorCondition to extractors with custom op class <class 'extensions.ops.TensorIterator_ops.TensorIteratorCondition'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry TensorIteratorBackEdge to extractors with custom op class <class 'extensions.ops.TensorIterator_ops.TensorIteratorBackEdge'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Accum to extractors with custom op class <class 'extensions.ops.accum.AccumOp'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Assert to extractors with custom op class <class 'extensions.ops.assert_op.Assert'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Axpy to extractors with custom op class <class 'extensions.ops.axpy.AxpyOp'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Binarization to extractors with custom op class <class 'extensions.ops.binarization.Binarization'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry BN to extractors with custom op class <class 'extensions.ops.bn.BNOp'>.
[ 2019-08-16 11:41:45,641 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry _contrib_box_nms to extractors with custom op class <class 'extensions.ops.box_nms.BoxNms'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry ConstantFill to extractors with custom op class <class 'extensions.ops.constant_fill.ConstantFill'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry copy to extractors with custom op class <class 'extensions.ops.copy.CopyOp'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Correlation to extractors with custom op class <class 'extensions.ops.correlation.CorrelationOp'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry DataAugmentation to extractors with custom op class <class 'extensions.ops.data_augmentation.DataAugmentationOp'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry ExperimentalDetectronDetectionOutput to extractors with custom op class <class 'extensions.ops.detectionoutput_onnx.ExperimentalDetectronDetectionOutput'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry FakeQuantize to extractors with custom op class <class 'extensions.ops.fakequantize.FakeQuantize'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry GRN to extractors with custom op class <class 'extensions.ops.grn.GRNOp'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry InstanceNormalization to extractors with custom op class <class 'extensions.ops.instance_normalization.InstanceNormalization'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Interp to extractors with custom op class <class 'extensions.ops.interp.InterpOp'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Interpolate to extractors with custom op class <class 'extensions.ops.interpolate.Interpolate'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry LSTMCell to extractors with custom op class <class 'extensions.ops.lstm_cell.LSTMCell'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry LSTMSequence to extractors with custom op class <class 'extensions.ops.lstm_sequence.LSTMSequence'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Merge to extractors with custom op class <class 'extensions.ops.merge.Merge'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry MVN to extractors with custom op class <class 'extensions.ops.mvn.MVN'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Normalize to extractors with custom op class <class 'extensions.ops.normalize.NormalizeOp'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Parameter to extractors with custom op class <class 'extensions.ops.parameter.Parameter'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry PowerFile to extractors with custom op class <class 'extensions.ops.power_file.PowerFileOp'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry PredictionHeatmap to extractors with custom op class <class 'extensions.ops.prediction_heatmap.PredictionHeatmapOp'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry PReLU to extractors with custom op class <class 'extensions.ops.prelu.PreluOp'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry PriorBox to extractors with custom op class <class 'extensions.ops.priorbox.PriorBoxOp'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry PriorBoxClustered to extractors with custom op class <class 'extensions.ops.priorbox_clustered.PriorBoxClusteredOp'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry ExperimentalDetectronPriorGridGenerator to extractors with custom op class <class 'extensions.ops.priorgridgenerator_onnx.ExperimentalDetectronPriorGridGenerator'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Proposal to extractors with custom op class <class 'extensions.ops.proposal.ProposalOp'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry ExperimentalDetectronGenerateProposalsSingleImage to extractors with custom op class <class 'extensions.ops.proposal_onnx.ExperimentalDetectronGenerateProposalsSingleImage'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry PSROIPooling to extractors with custom op class <class 'extensions.ops.psroipooling.PSROIPoolingOp'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Range to extractors with custom op class <class 'extensions.ops.range.Range'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry RegionYolo to extractors with custom op class <class 'extensions.ops.regionyolo.RegionYoloOp'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry ReorgYolo to extractors with custom op class <class 'extensions.ops.reorgyolo.ReorgYoloOp'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Resample to extractors with custom op class <class 'extensions.ops.resample.ResampleOp'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry ExperimentalDetectronROIFeatureExtractor to extractors with custom op class <class 'extensions.ops.roifeatureextractor_onnx.ExperimentalDetectronROIFeatureExtractor'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry ShuffleChannel to extractors with custom op class <class 'extensions.ops.shufflechannel.ShuffleChannelOp'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry SimplerNMS to extractors with custom op class <class 'extensions.ops.simplernms.SimplerNMSOp'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry SpatialTransformer to extractors with custom op class <class 'extensions.ops.spatial_transformer.SpatialTransformOp'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Splice to extractors with custom op class <class 'extensions.ops.splice.Splice'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry SwapAxis to extractors with custom op class <class 'extensions.ops.swapaxis.SwapAxis'>.
[ 2019-08-16 11:41:45,642 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Switch to extractors with custom op class <class 'extensions.ops.switch.Switch'>.
[ 2019-08-16 11:41:45,643 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry TensorIterator to extractors with custom op class <class 'extensions.ops.tensor_iterator.TensorIterator'>.
[ 2019-08-16 11:41:45,643 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry ExperimentalDetectronTopKROIs to extractors with custom op class <class 'extensions.ops.topkrois_onnx.ExperimentalDetectronTopKROIs'>.
[ 2019-08-16 11:41:45,643 ] [ DEBUG ] [ register_custom_ops:104 ]  Added a new entry Upsample to extractors with custom op class <class 'extensions.ops.upsample.UpsampleOp'>.
[ 2019-08-16 11:41:45,696 ] [ DEBUG ] [ utils:78 ]  value = [False], shape = [], res = [False], res.shape = (1,)
[ 2019-08-16 11:41:45,696 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,697 ] [ DEBUG ] [ utils:78 ]  value = [2], shape = [], res = [2], res.shape = (1,)
[ 2019-08-16 11:41:45,697 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,697 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,698 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,698 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,699 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,699 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ 2019-08-16 11:41:45,700 ] [ DEBUG ] [ utils:78 ]  value = [0.9999970197677612], shape = [], res = [0.999997], res.shape = (1,)
[ 2019-08-16 11:41:45,700 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,700 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ 2019-08-16 11:41:45,701 ] [ DEBUG ] [ utils:78 ]  value = [0.9999970197677612], shape = [], res = [0.999997], res.shape = (1,)
[ 2019-08-16 11:41:45,701 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,701 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [3] and value.shape = (3,)
[ 2019-08-16 11:41:45,702 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [3] and value.shape = (3,)
[ 2019-08-16 11:41:45,704 ] [ DEBUG ] [ utils:78 ]  value = [0.0010000000474974513], shape = [], res = [0.001], res.shape = (1,)
[ 2019-08-16 11:41:45,704 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,704 ] [ DEBUG ] [ utils:78 ]  value = [1.0], shape = [], res = [1.], res.shape = (1,)
[ 2019-08-16 11:41:45,705 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,705 ] [ DEBUG ] [ utils:78 ]  value = [1.0], shape = [], res = [1.], res.shape = (1,)
[ 2019-08-16 11:41:45,706 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,707 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ 2019-08-16 11:41:45,708 ] [ DEBUG ] [ utils:78 ]  value = [1.0], shape = [], res = [1.], res.shape = (1,)
[ 2019-08-16 11:41:45,708 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,709 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ 2019-08-16 11:41:45,710 ] [ DEBUG ] [ utils:78 ]  value = [0.0010000000474974513], shape = [], res = [0.001], res.shape = (1,)
[ 2019-08-16 11:41:45,710 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,712 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [ 3  3  1 16] and value.shape = (3, 3, 1, 16)
[ 2019-08-16 11:41:45,712 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [16] and value.shape = (16,)
[ 2019-08-16 11:41:45,713 ] [ DEBUG ] [ utils:78 ]  value = [0.20000000298023224], shape = [], res = [0.2], res.shape = (1,)
[ 2019-08-16 11:41:45,713 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,714 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [16] and value.shape = (16,)
[ 2019-08-16 11:41:45,715 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [16] and value.shape = (16,)
[ 2019-08-16 11:41:45,715 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [16] and value.shape = (16,)
[ 2019-08-16 11:41:45,715 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [16] and value.shape = (16,)
[ WARNING ]  Broadcast of scalar to shape: [0]
[ 2019-08-16 11:41:45,716 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [0] and value.shape = (0,)
[ WARNING ]  Broadcast of scalar to shape: [0]
[ 2019-08-16 11:41:45,716 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [0] and value.shape = (0,)
[ WARNING ]  FusedBatchNorm doesn't support is_training=True
[ 2019-08-16 11:41:45,718 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [ 3  3 16 32] and value.shape = (3, 3, 16, 32)
[ 2019-08-16 11:41:45,718 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [32] and value.shape = (32,)
[ 2019-08-16 11:41:45,719 ] [ DEBUG ] [ utils:78 ]  value = [0.20000000298023224], shape = [], res = [0.2], res.shape = (1,)
[ 2019-08-16 11:41:45,719 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,720 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [32] and value.shape = (32,)
[ 2019-08-16 11:41:45,721 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [32] and value.shape = (32,)
[ 2019-08-16 11:41:45,721 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [32] and value.shape = (32,)
[ 2019-08-16 11:41:45,721 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [32] and value.shape = (32,)
[ WARNING ]  Broadcast of scalar to shape: [0]
[ 2019-08-16 11:41:45,722 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [0] and value.shape = (0,)
[ WARNING ]  Broadcast of scalar to shape: [0]
[ 2019-08-16 11:41:45,723 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [0] and value.shape = (0,)
[ WARNING ]  FusedBatchNorm doesn't support is_training=True
[ 2019-08-16 11:41:45,724 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [ 1  7 32 96] and value.shape = (1, 7, 32, 96)
[ 2019-08-16 11:41:45,725 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [96] and value.shape = (96,)
[ 2019-08-16 11:41:45,726 ] [ DEBUG ] [ utils:78 ]  value = [0.20000000298023224], shape = [], res = [0.2], res.shape = (1,)
[ 2019-08-16 11:41:45,726 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,726 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [3] and value.shape = (3,)
[ 2019-08-16 11:41:45,727 ] [ DEBUG ] [ utils:78 ]  value = [3], shape = [], res = [3], res.shape = (1,)
[ 2019-08-16 11:41:45,727 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,727 ] [ DEBUG ] [ utils:78 ]  value = [2], shape = [], res = [2], res.shape = (1,)
[ 2019-08-16 11:41:45,727 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,727 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,727 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,728 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [2] and value.shape = (2,)
[ 2019-08-16 11:41:45,728 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,728 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,729 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,729 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,730 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ 2019-08-16 11:41:45,730 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,730 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,731 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ 2019-08-16 11:41:45,731 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,731 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,731 ] [ DEBUG ] [ utils:78 ]  value = [0.0], shape = [], res = [0.], res.shape = (1,)
[ 2019-08-16 11:41:45,732 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,732 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,733 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,733 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ 2019-08-16 11:41:45,733 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,734 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,734 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,735 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,735 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ 2019-08-16 11:41:45,736 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,736 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,736 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,736 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,736 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,737 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,737 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,737 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,741 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,741 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,742 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [192 192] and value.shape = (192, 192)
[ 2019-08-16 11:41:45,742 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [192] and value.shape = (192,)
[ 2019-08-16 11:41:45,742 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [192  96] and value.shape = (192, 96)
[ 2019-08-16 11:41:45,743 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [96] and value.shape = (96,)
[ 2019-08-16 11:41:45,743 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,743 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,744 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,744 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,745 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,745 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,746 ] [ DEBUG ] [ utils:78 ]  value = [1.0], shape = [], res = [1.], res.shape = (1,)
[ 2019-08-16 11:41:45,746 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,748 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,748 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,749 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,749 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,749 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,749 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,750 ] [ DEBUG ] [ utils:78 ]  value = [3], shape = [], res = [3], res.shape = (1,)
[ 2019-08-16 11:41:45,750 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,750 ] [ DEBUG ] [ utils:78 ]  value = [2], shape = [], res = [2], res.shape = (1,)
[ 2019-08-16 11:41:45,750 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,750 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,751 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,751 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [2] and value.shape = (2,)
[ 2019-08-16 11:41:45,751 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,751 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,752 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ 2019-08-16 11:41:45,752 ] [ DEBUG ] [ utils:78 ]  value = [3], shape = [], res = [3], res.shape = (1,)
[ 2019-08-16 11:41:45,752 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,753 ] [ DEBUG ] [ utils:78 ]  value = [2], shape = [], res = [2], res.shape = (1,)
[ 2019-08-16 11:41:45,753 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,753 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,753 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,753 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [2] and value.shape = (2,)
[ 2019-08-16 11:41:45,754 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,754 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,754 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,755 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,755 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ 2019-08-16 11:41:45,756 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,756 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,757 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ 2019-08-16 11:41:45,757 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,757 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,757 ] [ DEBUG ] [ utils:78 ]  value = [0.0], shape = [], res = [0.], res.shape = (1,)
[ 2019-08-16 11:41:45,757 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,758 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,758 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,759 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ 2019-08-16 11:41:45,759 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,759 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,760 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,760 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,761 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ 2019-08-16 11:41:45,761 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,761 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,761 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,761 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,762 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,762 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,763 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,763 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,766 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,766 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,767 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [192 192] and value.shape = (192, 192)
[ 2019-08-16 11:41:45,767 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [192] and value.shape = (192,)
[ 2019-08-16 11:41:45,768 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [192  96] and value.shape = (192, 96)
[ 2019-08-16 11:41:45,768 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [96] and value.shape = (96,)
[ 2019-08-16 11:41:45,769 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,769 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,770 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,770 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,770 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,771 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,772 ] [ DEBUG ] [ utils:78 ]  value = [1.0], shape = [], res = [1.], res.shape = (1,)
[ 2019-08-16 11:41:45,772 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,773 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,773 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,774 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,774 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,775 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,775 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,775 ] [ DEBUG ] [ utils:78 ]  value = [3], shape = [], res = [3], res.shape = (1,)
[ 2019-08-16 11:41:45,775 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,776 ] [ DEBUG ] [ utils:78 ]  value = [2], shape = [], res = [2], res.shape = (1,)
[ 2019-08-16 11:41:45,776 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,776 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,776 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,776 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [2] and value.shape = (2,)
[ 2019-08-16 11:41:45,777 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,777 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,777 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ 2019-08-16 11:41:45,778 ] [ DEBUG ] [ utils:78 ]  value = [2], shape = [], res = [2], res.shape = (1,)
[ 2019-08-16 11:41:45,778 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,778 ] [ DEBUG ] [ utils:78 ]  value = [3], shape = [], res = [3], res.shape = (1,)
[ 2019-08-16 11:41:45,778 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,778 ] [ DEBUG ] [ utils:78 ]  value = [2], shape = [], res = [2], res.shape = (1,)
[ 2019-08-16 11:41:45,778 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,779 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,779 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,779 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [2] and value.shape = (2,)
[ 2019-08-16 11:41:45,779 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,779 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,780 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,781 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,781 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ 2019-08-16 11:41:45,781 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,782 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,782 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ 2019-08-16 11:41:45,782 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,782 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,783 ] [ DEBUG ] [ utils:78 ]  value = [0.0], shape = [], res = [0.], res.shape = (1,)
[ 2019-08-16 11:41:45,783 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,783 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,784 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,784 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ 2019-08-16 11:41:45,785 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,785 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,785 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,786 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,786 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ 2019-08-16 11:41:45,787 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,787 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,787 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,787 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,788 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,788 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,788 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,788 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,792 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,792 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,793 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [288 192] and value.shape = (288, 192)
[ 2019-08-16 11:41:45,793 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [192] and value.shape = (192,)
[ 2019-08-16 11:41:45,794 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [288  96] and value.shape = (288, 96)
[ 2019-08-16 11:41:45,794 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [96] and value.shape = (96,)
[ 2019-08-16 11:41:45,794 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,794 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,796 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,796 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,796 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,796 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,798 ] [ DEBUG ] [ utils:78 ]  value = [1.0], shape = [], res = [1.], res.shape = (1,)
[ 2019-08-16 11:41:45,798 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,799 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,799 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,800 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,800 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,800 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,800 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,801 ] [ DEBUG ] [ utils:78 ]  value = [3], shape = [], res = [3], res.shape = (1,)
[ 2019-08-16 11:41:45,801 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,801 ] [ DEBUG ] [ utils:78 ]  value = [2], shape = [], res = [2], res.shape = (1,)
[ 2019-08-16 11:41:45,801 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,802 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,802 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,802 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [2] and value.shape = (2,)
[ 2019-08-16 11:41:45,802 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,802 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,803 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ 2019-08-16 11:41:45,803 ] [ DEBUG ] [ utils:78 ]  value = [3], shape = [], res = [3], res.shape = (1,)
[ 2019-08-16 11:41:45,804 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,804 ] [ DEBUG ] [ utils:78 ]  value = [2], shape = [], res = [2], res.shape = (1,)
[ 2019-08-16 11:41:45,804 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,804 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,804 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,805 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [2] and value.shape = (2,)
[ 2019-08-16 11:41:45,805 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,805 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,806 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,806 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,806 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ 2019-08-16 11:41:45,807 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,807 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,807 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ 2019-08-16 11:41:45,808 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,808 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,808 ] [ DEBUG ] [ utils:78 ]  value = [0.0], shape = [], res = [0.], res.shape = (1,)
[ 2019-08-16 11:41:45,808 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,809 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,809 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,809 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ 2019-08-16 11:41:45,810 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,810 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,811 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,811 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,812 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ 2019-08-16 11:41:45,812 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,812 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,812 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,812 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,813 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,813 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,814 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,814 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,817 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,817 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,818 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [288 192] and value.shape = (288, 192)
[ 2019-08-16 11:41:45,818 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [192] and value.shape = (192,)
[ 2019-08-16 11:41:45,819 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [288  96] and value.shape = (288, 96)
[ 2019-08-16 11:41:45,819 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [96] and value.shape = (96,)
[ 2019-08-16 11:41:45,820 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,820 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,821 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,821 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,821 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,821 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,823 ] [ DEBUG ] [ utils:78 ]  value = [1.0], shape = [], res = [1.], res.shape = (1,)
[ 2019-08-16 11:41:45,823 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,824 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,824 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,826 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,826 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,826 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,826 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,827 ] [ DEBUG ] [ utils:78 ]  value = [3], shape = [], res = [3], res.shape = (1,)
[ 2019-08-16 11:41:45,827 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,827 ] [ DEBUG ] [ utils:78 ]  value = [2], shape = [], res = [2], res.shape = (1,)
[ 2019-08-16 11:41:45,827 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,827 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,827 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,828 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [2] and value.shape = (2,)
[ 2019-08-16 11:41:45,828 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,828 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,829 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ 2019-08-16 11:41:45,829 ] [ DEBUG ] [ utils:78 ]  value = [2], shape = [], res = [2], res.shape = (1,)
[ 2019-08-16 11:41:45,829 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,829 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [4] and value.shape = (4,)
[ 2019-08-16 11:41:45,830 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [  1   1 192  41] and value.shape = (1, 1, 192, 41)
[ 2019-08-16 11:41:45,830 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [41] and value.shape = (41,)
[ 2019-08-16 11:41:45,831 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [3] and value.shape = (3,)
[ 2019-08-16 11:41:45,832 ] [ DEBUG ] [ utils:78 ]  value = [2], shape = [], res = [2], res.shape = (1,)
[ 2019-08-16 11:41:45,832 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,832 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,832 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,833 ] [ DEBUG ] [ utils:78 ]  value = [3], shape = [], res = [3], res.shape = (1,)
[ 2019-08-16 11:41:45,833 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,833 ] [ DEBUG ] [ utils:78 ]  value = [1], shape = [], res = [1], res.shape = (1,)
[ 2019-08-16 11:41:45,834 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,834 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,835 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ 2019-08-16 11:41:45,835 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,835 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,836 ] [ DEBUG ] [ utils:78 ]  value = [-1], shape = [], res = [-1], res.shape = (1,)
[ 2019-08-16 11:41:45,836 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,836 ] [ DEBUG ] [ utils:78 ]  value = [-1], shape = [], res = [-1], res.shape = (1,)
[ 2019-08-16 11:41:45,836 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,837 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [3] and value.shape = (3,)
[ 2019-08-16 11:41:45,838 ] [ DEBUG ] [ utils:78 ]  value = [0], shape = [], res = [0], res.shape = (1,)
[ 2019-08-16 11:41:45,838 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,838 ] [ DEBUG ] [ utils:78 ]  value = [112], shape = [], res = [112], res.shape = (1,)
[ 2019-08-16 11:41:45,838 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,838 ] [ DEBUG ] [ utils:78 ]  value = [28], shape = [], res = [28], res.shape = (1,)
[ 2019-08-16 11:41:45,838 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,839 ] [ DEBUG ] [ utils:78 ]  value = [28], shape = [], res = [28], res.shape = (1,)
[ 2019-08-16 11:41:45,839 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,839 ] [ DEBUG ] [ utils:78 ]  value = [41], shape = [], res = [41], res.shape = (1,)
[ 2019-08-16 11:41:45,839 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ WARNING ]  Broadcast of scalar to shape: [40]
[ 2019-08-16 11:41:45,839 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [40] and value.shape = (40,)
[ 2019-08-16 11:41:45,840 ] [ DEBUG ] [ utils:78 ]  value = [0.5], shape = [], res = [0.5], res.shape = (1,)
[ 2019-08-16 11:41:45,840 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,840 ] [ DEBUG ] [ utils:78 ]  value = [5.0], shape = [], res = [5.], res.shape = (1,)
[ 2019-08-16 11:41:45,840 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,841 ] [ DEBUG ] [ utils:78 ]  value = [1.0], shape = [], res = [1.], res.shape = (1,)
[ 2019-08-16 11:41:45,841 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,841 ] [ DEBUG ] [ utils:78 ]  value = [0.5], shape = [], res = [0.5], res.shape = (1,)
[ 2019-08-16 11:41:45,842 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,842 ] [ DEBUG ] [ utils:78 ]  value = [-1], shape = [], res = [-1], res.shape = (1,)
[ 2019-08-16 11:41:45,842 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [] and value.shape = ()
[ 2019-08-16 11:41:45,843 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [2] and value.shape = (2,)
[ 2019-08-16 11:41:45,843 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [2] and value.shape = (2,)
[ 2019-08-16 11:41:45,844 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [4] and value.shape = (4,)
[ 2019-08-16 11:41:45,844 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [  1   1   1 128] and value.shape = (1, 1, 1, 128)
[ 2019-08-16 11:41:45,844 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [128] and value.shape = (128,)
[ 2019-08-16 11:41:45,845 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [2] and value.shape = (2,)
[ 2019-08-16 11:41:45,846 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [4] and value.shape = (4,)
[ 2019-08-16 11:41:45,846 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [  1   1 128 128] and value.shape = (1, 1, 128, 128)
[ 2019-08-16 11:41:45,847 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [128] and value.shape = (128,)
[ 2019-08-16 11:41:45,848 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [2] and value.shape = (2,)
[ 2019-08-16 11:41:45,848 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [4] and value.shape = (4,)
[ 2019-08-16 11:41:45,848 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [  1   1 128 128] and value.shape = (1, 1, 128, 128)
[ 2019-08-16 11:41:45,849 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [128] and value.shape = (128,)
[ 2019-08-16 11:41:45,850 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [2] and value.shape = (2,)
[ 2019-08-16 11:41:45,850 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [4] and value.shape = (4,)
[ 2019-08-16 11:41:45,851 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [  1   1 128   1] and value.shape = (1, 1, 128, 1)
[ WARNING ]  Broadcast of scalar to shape: [1]
[ 2019-08-16 11:41:45,851 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [1] and value.shape = (1,)
[ 2019-08-16 11:41:45,852 ] [ DEBUG ] [ const:33 ]  Constant extractor for node gives shape = [2] and value.shape = (2,)
[ WARNING ]  Instructions/layers that do not have attribute extractors:
[ WARNING ]      Abs (1)
[ WARNING ]          Abs
[ WARNING ]      Sign (1)
[ WARNING ]          Sign
[ WARNING ]      SparseToDense (1)
[ WARNING ]          predictions/decoded
[ 2019-08-16 11:41:45,853 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.ExpandDimsToUnsqueeze.ExpandDimsToUnsqueeze'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,853 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.ExpandDimsToUnsqueeze.ExpandDimsToUnsqueeze'> will be run after <class 'extensions.front.Pack.Pack'>
[ 2019-08-16 11:41:45,853 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.FillToBroadcast.FillToBroadcast'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,853 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.FillToBroadcast.FillToBroadcast'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,853 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.SqueezeNormalize.SqueezeNormalize'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,854 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.SqueezeNormalize.SqueezeNormalize'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,854 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.TopKNormalize.TopKNormalize'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,854 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.TopKNormalize.TopKNormalize'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,854 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.binary_quantize_normalization.BinaryFakeQuantizeNormalization'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,854 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.binary_quantize_normalization.BinaryFakeQuantizeNormalization'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,854 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.create_tensor_nodes.CreateTensorNodes'> will be run after <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,854 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.disable_weights_quantize_value_propagation.DisableQuantizeValuePropagation'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,854 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.disable_weights_quantize_value_propagation.DisableQuantizeValuePropagation'> will be run after <class 'extensions.front.tf.FakeQuantWithMinMaxVars.FakeQuantWithMinMaxVarsToQuantize'>
[ 2019-08-16 11:41:45,854 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.input_cut.InputCut'> will be run after <class 'extensions.front.output_cut.OutputCut'>
[ 2019-08-16 11:41:45,854 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.output_cut.OutputCut'> will be run after <class 'extensions.front.user_data_repack.UserDataRepack'>
[ 2019-08-16 11:41:45,854 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.override_batch.OverrideBatch'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,854 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.override_batch.OverrideBatch'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,854 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.GNMT_DynamicSequenceLengths.GNMT_sequence_lengths'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,854 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.GNMT_DynamicSequenceLengths.GNMT_sequence_lengths'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,854 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.tensorflow_custom_operations_config_update.TensorflowCustomOperationsConfigUpdate'> will be run after <class 'extensions.front.freeze_placeholder_value.FreezePlaceholderValue'>
[ 2019-08-16 11:41:45,854 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.tensorflow_patterns.TensorflowSubgraphPatterns'> will be run after <class 'extensions.front.tf.tensorflow_custom_operations_config_update.TensorflowCustomOperationsConfigUpdate'>
[ 2019-08-16 11:41:45,854 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.tensorflow_patterns.TensorflowOperationPatterns'> will be run before <class 'extensions.front.tf.tensorflow_use_custom_operations_config.TensorflowUseCustomOperationsConfig'>
[ 2019-08-16 11:41:45,854 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.tensorflow_patterns.TensorflowOperationPatterns'> will be run after <class 'extensions.front.tf.tensorflow_patterns.TensorflowSubgraphPatterns'>
[ 2019-08-16 11:41:45,854 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.tensorflow_use_custom_operations_config.TensorflowUseCustomOperationsConfig'> will be run before <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,854 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.variables_values_freezing.VariablesToConstants'> will be run before <class 'extensions.front.freeze_placeholder_value.FreezePlaceholderValue'>
[ 2019-08-16 11:41:45,854 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.variables_values_freezing.VariablesToConstants'> will be run after <class 'extensions.front.input_cut.InputCut'>
[ 2019-08-16 11:41:45,854 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIPreprocessorReplacement'> will be run before <class 'extensions.front.Pack.Pack'>
[ 2019-08-16 11:41:45,854 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIPreprocessorReplacement'> will be run before <class 'extensions.front.sub.Sub'>
[ 2019-08-16 11:41:45,854 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIPreprocessorReplacement'> will be run before <class 'extensions.front.TransposeOrderNormalizer.TransposeOrderNormalizer'>
[ 2019-08-16 11:41:45,854 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIPreprocessorReplacement'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,854 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIDetectionOutputReplacement'> will be run before <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIMaskRCNNROIPoolingSecondReplacement'>
[ 2019-08-16 11:41:45,855 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIDetectionOutputReplacement'> will be run before <class 'extensions.front.tf.Unpack.Unpack'>
[ 2019-08-16 11:41:45,855 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIDetectionOutputReplacement'> will be run before <class 'extensions.front.sub.Sub'>
[ 2019-08-16 11:41:45,855 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIDetectionOutputReplacement'> will be run before <class 'extensions.front.TransposeOrderNormalizer.TransposeOrderNormalizer'>
[ 2019-08-16 11:41:45,855 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIDetectionOutputReplacement'> will be run after <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIProposalReplacement'>
[ 2019-08-16 11:41:45,855 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIDetectionOutputReplacement'> will be run after <class 'extensions.front.tf.CropAndResizeReplacement.CropAndResizeReplacement'>
[ 2019-08-16 11:41:45,855 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIMaskRCNNROIPoolingSecondReplacement'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,855 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIMaskRCNNROIPoolingSecondReplacement'> will be run after <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIProposalReplacement'>
[ 2019-08-16 11:41:45,855 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIProposalReplacement'> will be run before <class 'extensions.front.sub.Sub'>
[ 2019-08-16 11:41:45,855 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIProposalReplacement'> will be run before <class 'extensions.front.tf.CropAndResizeReplacement.CropAndResizeReplacement'>
[ 2019-08-16 11:41:45,855 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIProposalReplacement'> will be run before <class 'extensions.front.TransposeOrderNormalizer.TransposeOrderNormalizer'>
[ 2019-08-16 11:41:45,855 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIProposalReplacement'> will be run after <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIPreprocessorReplacement'>
[ 2019-08-16 11:41:45,855 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPISSDPostprocessorReplacement'> will be run before <class 'extensions.front.div.Div'>
[ 2019-08-16 11:41:45,855 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPISSDPostprocessorReplacement'> will be run before <class 'extensions.front.standalone_const_eraser.StandaloneConstEraser'>
[ 2019-08-16 11:41:45,855 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPISSDPostprocessorReplacement'> will be run before <class 'extensions.front.TransposeOrderNormalizer.TransposeOrderNormalizer'>
[ 2019-08-16 11:41:45,855 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPISSDPostprocessorReplacement'> will be run after <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIPreprocessorReplacement'>
[ 2019-08-16 11:41:45,855 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIPSROIPoolingReplacement'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,855 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIPSROIPoolingReplacement'> will be run after <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIProposalReplacement'>
[ 2019-08-16 11:41:45,855 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIPSROIPoolingReplacement'> will be run after <class 'extensions.front.TransposeOrderNormalizer.TransposeOrderNormalizer'>
[ 2019-08-16 11:41:45,855 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.RetinaNetFilteredDetectionsReplacement.RetinaNetFilteredDetectionsReplacement'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,855 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.RetinaNetFilteredDetectionsReplacement.RetinaNetFilteredDetectionsReplacement'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,855 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.SSDToolboxDetectionOutput.SSDToolboxDetectionOutputReplacement'> will be run before <class 'extensions.front.standalone_const_eraser.StandaloneConstEraser'>
[ 2019-08-16 11:41:45,855 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.SSDToolboxDetectionOutput.SSDToolboxDetectionOutputReplacement'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,855 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIMaskRCNNSigmoidReplacement'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,855 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIMaskRCNNSigmoidReplacement'> will be run after <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIMaskRCNNROIPoolingSecondReplacement'>
[ 2019-08-16 11:41:45,855 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIOutputReplacement'> will be run before <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIPreprocessorReplacement'>
[ 2019-08-16 11:41:45,855 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIOutputReplacement'> will be run before <class 'extensions.front.TransposeOrderNormalizer.TransposeOrderNormalizer'>
[ 2019-08-16 11:41:45,856 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIOutputReplacement'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,856 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIConstValueOverride'> will be run before <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIPreprocessorReplacement'>
[ 2019-08-16 11:41:45,856 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIConstValueOverride'> will be run before <class 'extensions.front.TransposeOrderNormalizer.TransposeOrderNormalizer'>
[ 2019-08-16 11:41:45,856 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIConstValueOverride'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,856 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.YOLO.YoloRegionAddon'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,856 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.YOLO.YoloRegionAddon'> will be run after <class 'extensions.front.no_op_eraser.NoOpEraser'>
[ 2019-08-16 11:41:45,856 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.YOLO.YoloRegionAddon'> will be run after <class 'extensions.front.standalone_const_eraser.StandaloneConstEraser'>
[ 2019-08-16 11:41:45,856 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.YOLO.YoloV3RegionAddon'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,856 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.YOLO.YoloV3RegionAddon'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,856 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.InterpolateNormalizer.InterpolateNormalizer'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,856 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.InterpolateNormalizer.InterpolateNormalizer'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,856 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.LRNReplacer.LRNReplacer'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,856 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.LRNReplacer.LRNReplacer'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,856 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.Pack.Pack'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,856 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.Pack.Pack'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,856 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.PowerToEltwises.PowerToEltwises'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,856 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.PowerToEltwises.PowerToEltwises'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,856 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.FakeQuantWithMinMaxVars.FakeQuantWithMinMaxVarsToQuantize'> will be run before <class 'extensions.front.sub.Sub'>
[ 2019-08-16 11:41:45,856 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.FakeQuantWithMinMaxVars.FakeQuantWithMinMaxVarsToQuantize'> will be run before <class 'extensions.front.div.Div'>
[ 2019-08-16 11:41:45,856 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.FakeQuantWithMinMaxVars.FakeQuantWithMinMaxVarsToQuantize'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,856 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.div.Div'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,856 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.div.Div'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,856 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.eltwise_n.EltwiseNReplacement'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,856 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.eltwise_n.EltwiseNReplacement'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,856 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.image_scaler.ImageScaler'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,856 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.image_scaler.ImageScaler'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,856 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.instance_normalization.InstanceNormalization'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,857 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.instance_normalization.InstanceNormalization'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,857 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.reciprocal.ReciprocalReplacer'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,857 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.reciprocal.ReciprocalReplacer'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,857 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.sub.Sub'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,857 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.sub.Sub'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,857 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.squared_difference.SquaredDifference'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,857 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.squared_difference.SquaredDifference'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,857 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.BlockLSTM.BlockLSTM'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,857 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.BlockLSTM.BlockLSTM'> will be run after <class 'extensions.front.restore_ports.RestorePorts'>
[ 2019-08-16 11:41:45,857 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.CropAndResizeReplacement.CropAndResizeReplacement'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,857 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.CropAndResizeReplacement.CropAndResizeReplacement'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,857 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.LogSoftmax.LogSoftmaxFrontReplacer'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,857 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.LogSoftmax.LogSoftmaxFrontReplacer'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,857 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.Unpack.Unpack'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,857 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.Unpack.Unpack'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,857 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.SizeReplacer.SizeFrontReplacer'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,857 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.SizeReplacer.SizeFrontReplacer'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,857 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.assign_elimination.AssignElimination'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,857 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.assign_elimination.AssignElimination'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,857 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.assign_elimination.AssignSubElimination'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,857 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.assign_elimination.AssignSubElimination'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,857 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.assign_elimination.AssignAddElimination'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,857 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.assign_elimination.AssignAddElimination'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,857 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.assign_elimination.AssertElimination'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,857 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.assign_elimination.AssertElimination'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,857 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.fake_const_ext.FakeConstToConst'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,857 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.fake_const_ext.FakeConstToConst'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,858 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.TransposeOrderNormalizer.TransposeOrderNormalizer'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,858 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.TransposeOrderNormalizer.TransposeOrderNormalizer'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,858 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.freeze_placeholder_value.FreezePlaceholderValue'> will be run after <class 'extensions.front.restore_ports.RestorePorts'>
[ 2019-08-16 11:41:45,858 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.no_op_eraser.NoOpEraser'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,858 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.no_op_eraser.NoOpEraser'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,858 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.reduce_axis_normalizer.ReduceAxisNormalizer'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,858 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.reduce_axis_normalizer.ReduceAxisNormalizer'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,858 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.reshape_dim_normalizer.ReshapeDimNormalizer'> will be run before <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,858 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.reshape_dim_normalizer.ReshapeDimNormalizer'> will be run after <class 'extensions.front.freeze_placeholder_value.FreezePlaceholderValue'>
[ 2019-08-16 11:41:45,858 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.restore_ports.RestorePorts'> will be run after <class 'extensions.front.input_cut.InputCut'>
[ 2019-08-16 11:41:45,858 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.softmax.SoftmaxFromKeras'> will be run before <class 'extensions.front.sub.Sub'>
[ 2019-08-16 11:41:45,858 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.softmax.SoftmaxFromKeras'> will be run before <class 'extensions.front.div.Div'>
[ 2019-08-16 11:41:45,858 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.softmax.SoftmaxFromKeras'> will be run after <class 'extensions.front.reduce_axis_normalizer.ReduceAxisNormalizer'>
[ 2019-08-16 11:41:45,858 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.standalone_const_eraser.StandaloneConstEraser'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,858 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.standalone_const_eraser.StandaloneConstEraser'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,858 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.CTCGreedyDecoder.CTCGreedyDecoderReplacement'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,858 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.CTCGreedyDecoder.CTCGreedyDecoderReplacement'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,858 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.nearest_neighbor_upsampling.NearestNeighborUpsampling'> will be run before <class 'extensions.front.Pack.Pack'>
[ 2019-08-16 11:41:45,858 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.nearest_neighbor_upsampling.NearestNeighborUpsampling'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,858 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.FlattenToReshape.FlattenToReshapeableReshape'> will be run before <class 'extensions.front.Pack.Pack'>
[ 2019-08-16 11:41:45,858 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.FlattenToReshape.FlattenToReshapeableReshape'> will be run after <class 'extensions.front.tf.nearest_neighbor_upsampling.NearestNeighborUpsampling'>
[ 2019-08-16 11:41:45,858 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.basic_lstm_cell.BasicLSTMCell'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,858 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.basic_lstm_cell.BasicLSTMCell'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,858 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.concat.Concat'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,858 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.concat.Concat'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,858 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.fifo_replacer.FIFOQueue'> will be run before <class 'extensions.front.override_batch.OverrideBatch'>
[ 2019-08-16 11:41:45,859 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.fifo_replacer.FIFOQueue'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,859 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.fifo_replacer.QueueDequeueManyV2'> will be run before <class 'extensions.front.override_batch.OverrideBatch'>
[ 2019-08-16 11:41:45,859 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.fifo_replacer.QueueDequeueManyV2'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,859 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.mvn.MVN'> will be run before <class 'extensions.front.squared_difference.SquaredDifference'>
[ 2019-08-16 11:41:45,859 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.mvn.MVN'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,859 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.mvn_unrolled.MVNUnrolled'> will be run before <class 'extensions.front.squared_difference.SquaredDifference'>
[ 2019-08-16 11:41:45,859 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.mvn_unrolled.MVNUnrolled'> will be run before <class 'extensions.front.div.Div'>
[ 2019-08-16 11:41:45,859 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.mvn_unrolled.MVNUnrolled'> will be run before <class 'extensions.front.sub.Sub'>
[ 2019-08-16 11:41:45,859 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.mvn_unrolled.MVNUnrolled'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,859 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.prelu.PReLU'> will be run before <class 'extensions.front.PowerToEltwises.PowerToEltwises'>
[ 2019-08-16 11:41:45,859 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.prelu.PReLU'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,859 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.tf.prelu.PReLUWithAbs'> will be run before <class 'extensions.front.sub.Sub'>
[ 2019-08-16 11:41:45,859 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.tf.prelu.PReLUWithAbs'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,859 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.front.ArgMaxSqueeze.ArgMaxSqueeze'> will be run before <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:45,859 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.front.ArgMaxSqueeze.ArgMaxSqueeze'> will be run after <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:45,863 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.user_data_repack.UserDataRepack'>
[ 2019-08-16 11:41:45,869 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.output_cut.OutputCut'>
[ 2019-08-16 11:41:45,870 ] [ DEBUG ] [ graph:947 ]  Sink: seq_len/sink_port_0 for node seq_len
[ 2019-08-16 11:41:45,870 ] [ DEBUG ] [ graph:948 ]  {'precision': 'FP32', 'kind': 'op', 'op': 'Result', 'type': 'Result', 'infer': <function Result.__init__.<locals>.<lambda> at 0x7fe50a261c80>, 'value': None, 'data_type': None, 'in_ports_count': 1, 'name': 'seq_len/sink_port_0', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a266b70>), 'name', 'precision', 'type'], [('data', [], []), '@ports', '@consts'])], '_in_ports': {0: {}}, '_out_ports': {}}
[ 2019-08-16 11:41:45,870 ] [ DEBUG ] [ graph:949 ]  Add edge from seq_len to seq_len/sink_port_0
[ 2019-08-16 11:41:45,871 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
seq_len/sink_port_0
[ 2019-08-16 11:41:45,907 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: Reshape_4
Reshape_4/shape
conv2d/bias_1
conv2d/bias_1/read
conv2d/kernel_1
conv2d/kernel_1/read
conv2d_1/bias
conv2d_1/bias/read
conv2d_1/kernel
conv2d_1/kernel/read
conv2d_2/bias
conv2d_2/bias/read
conv2d_2/kernel
conv2d_2/kernel/read
conv2d_3/bias
conv2d_3/bias/read
conv2d_3/kernel
conv2d_3/kernel/read
metrics/Prod
metrics/Prod/reduction_indices
metrics/Reshape
metrics/Reshape/shape
metrics/Reshape_1
metrics/Reshape_1/shape
metrics/Reshape_2
metrics/Reshape_2/shape
metrics/Reshape_3
metrics/Reshape_3/shape
metrics/Reshape_4
metrics/Reshape_4/shape
metrics/Reshape_5
metrics/Reshape_5/shape
metrics/Reshape_6
metrics/Reshape_6/shape
metrics/Sqrt
metrics/Tanh
metrics/add
metrics/add/x
metrics/confidence
metrics/confidence/shape
metrics/conv2d/BiasAdd
metrics/conv2d/Conv2D
metrics/conv2d/Relu
metrics/conv2d_1/BiasAdd
metrics/conv2d_1/Conv2D
metrics/conv2d_1/Relu
metrics/conv2d_2/BiasAdd
metrics/conv2d_2/Conv2D
metrics/conv2d_2/Relu
metrics/conv2d_3/BiasAdd
metrics/conv2d_3/Conv2D
metrics/mul
metrics/mul/x
metrics/mul_1
metrics/mul_1/x
metrics/reliability
metrics/reliability/shape
metrics/sub
metrics/sub/y
predictions/CTCGreedyDecoder
predictions/Cast
predictions/Max
predictions/Max/reduction_indices
predictions/Rank
predictions/Reshape
predictions/Shape
predictions/Shape_1
predictions/Slice
predictions/Slice/begin
predictions/Slice/size
predictions/Softmax
predictions/Sub
predictions/Sub/y
predictions/activations
predictions/argmax
predictions/argmax/dimension
predictions/concat
predictions/concat/axis
predictions/concat/values_0
predictions/decoded
predictions/decoded/default_value
predictions/transpose
predictions/transpose/perm
shapes/img_h
shapes/img_w
shapes/letters
shapes/pred_h
shapes/pred_w
[ 2019-08-16 11:41:45,915 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.input_cut.InputCut'>
[ 2019-08-16 11:41:45,921 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
seq_len/sink_port_0
[ 2019-08-16 11:41:45,954 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 
[ INFO ]  Skip replacer <class 'extensions.front.tf.variables_values_freezing.VariablesToConstants'> (graph_condition not satisfied)
[ 2019-08-16 11:41:45,962 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.restore_ports.RestorePorts'>
[ 2019-08-16 11:41:45,969 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.freeze_placeholder_value.FreezePlaceholderValue'>
[ INFO ]  Skip replacer <class 'extensions.front.tf.tensorflow_custom_operations_config_update.TensorflowCustomOperationsConfigUpdate'> (graph_condition not satisfied)
[ INFO ]  Skip replacer <class 'extensions.front.tf.tensorflow_patterns.TensorflowSubgraphPatterns'> (graph_condition not satisfied)
[ INFO ]  Skip replacer <class 'extensions.front.tf.tensorflow_patterns.TensorflowOperationPatterns'> (graph_condition not satisfied)
[ INFO ]  Skip replacer <class 'extensions.front.tf.tensorflow_use_custom_operations_config.TensorflowUseCustomOperationsConfig'> (graph_condition not satisfied)
[ 2019-08-16 11:41:45,988 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.reshape_dim_normalizer.ReshapeDimNormalizer'>
[ 2019-08-16 11:41:46,016 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.pass_separator.FrontStart'>
[ 2019-08-16 11:41:46,022 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.YOLO.YoloV3RegionAddon'>
[ WARNING ]  Configuration file for custom replacement with id 'TFYOLOV3' doesn't exist
[ INFO ]  Failed to find custom replacement description with id 'TFYOLOV3'
[ 2019-08-16 11:41:46,028 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.TopKNormalize.TopKNormalize'>
[ 2019-08-16 11:41:46,034 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.SqueezeNormalize.SqueezeNormalize'>
[ 2019-08-16 11:41:46,041 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.SizeReplacer.SizeFrontReplacer'>
[ 2019-08-16 11:41:46,059 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.SSDToolboxDetectionOutput.SSDToolboxDetectionOutputReplacement'>
[ WARNING ]  Configuration file for custom replacement with id 'SSDToolboxDetectionOutput' doesn't exist
[ INFO ]  Failed to find custom replacement description with id 'SSDToolboxDetectionOutput'
[ 2019-08-16 11:41:46,065 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.RetinaNetFilteredDetectionsReplacement.RetinaNetFilteredDetectionsReplacement'>
[ WARNING ]  Configuration file for custom replacement with id 'RetinaNetFilteredDetectionsReplacement' doesn't exist
[ INFO ]  Failed to find custom replacement description with id 'RetinaNetFilteredDetectionsReplacement'
[ 2019-08-16 11:41:46,071 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.reduce_axis_normalizer.ReduceAxisNormalizer'>
[ 2019-08-16 11:41:46,100 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.softmax.SoftmaxFromKeras'>
[ 2019-08-16 11:41:46,123 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.reciprocal.ReciprocalReplacer'>
[ 2019-08-16 11:41:46,142 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.fifo_replacer.QueueDequeueManyV2'>
[ 2019-08-16 11:41:46,160 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.prelu.PReLUWithAbs'>
[ 2019-08-16 11:41:46,182 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.prelu.PReLU'>
[ 2019-08-16 11:41:46,208 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.PowerToEltwises.PowerToEltwises'>
[ 2019-08-16 11:41:46,225 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
seq_len/sink_port_0
[ 2019-08-16 11:41:46,258 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: batch_normalization/Sqrt
batch_normalization/batchnorm/Rsqrt
[ 2019-08-16 11:41:46,266 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIOutputReplacement'>
[ WARNING ]  Configuration file for custom replacement with id 'ObjectDetectionAPIOutputReplacement' doesn't exist
[ INFO ]  Failed to find custom replacement description with id 'ObjectDetectionAPIOutputReplacement'
[ 2019-08-16 11:41:46,272 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIConstValueOverride'>
[ WARNING ]  Configuration file for custom replacement with id 'ObjectDetectionAPIConstValueOverride' doesn't exist
[ INFO ]  Failed to find custom replacement description with id 'ObjectDetectionAPIConstValueOverride'
[ 2019-08-16 11:41:46,278 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIPreprocessorReplacement'>
[ WARNING ]  Configuration file for custom replacement with id 'ObjectDetectionAPIPreprocessorReplacement' doesn't exist
[ INFO ]  Failed to find custom replacement description with id 'ObjectDetectionAPIPreprocessorReplacement'
[ 2019-08-16 11:41:46,284 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPISSDPostprocessorReplacement'>
[ WARNING ]  Configuration file for custom replacement with id 'ObjectDetectionAPISSDPostprocessorReplacement' doesn't exist
[ INFO ]  Failed to find custom replacement description with id 'ObjectDetectionAPISSDPostprocessorReplacement'
[ 2019-08-16 11:41:46,289 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.standalone_const_eraser.StandaloneConstEraser'>
[ 2019-08-16 11:41:46,318 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIProposalReplacement'>
[ WARNING ]  Configuration file for custom replacement with id 'ObjectDetectionAPIProposalReplacement' doesn't exist
[ INFO ]  Failed to find custom replacement description with id 'ObjectDetectionAPIProposalReplacement'
[ 2019-08-16 11:41:46,324 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.no_op_eraser.NoOpEraser'>
[ 2019-08-16 11:41:46,343 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.YOLO.YoloRegionAddon'>
[ WARNING ]  Configuration file for custom replacement with id 'TFYOLO' doesn't exist
[ INFO ]  Failed to find custom replacement description with id 'TFYOLO'
[ 2019-08-16 11:41:46,349 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.nearest_neighbor_upsampling.NearestNeighborUpsampling'>
[ 2019-08-16 11:41:46,376 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.FlattenToReshape.FlattenToReshapeableReshape'>
[ 2019-08-16 11:41:46,397 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.Pack.Pack'>
[ 2019-08-16 11:41:46,415 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.ExpandDimsToUnsqueeze.ExpandDimsToUnsqueeze'>
[ 2019-08-16 11:41:46,422 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.mvn_unrolled.MVNUnrolled'>
[ 2019-08-16 11:41:46,422 ] [ DEBUG ] [ mvn_unrolled:34 ]  Enabled MVN replacement
[ 2019-08-16 11:41:46,439 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.mvn.MVN'>
[ 2019-08-16 11:41:46,440 ] [ DEBUG ] [ mvn:33 ]  Enabled MVN replacement
[ 2019-08-16 11:41:46,457 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.squared_difference.SquaredDifference'>
[ 2019-08-16 11:41:46,473 ] [ DEBUG ] [ replacement:114 ]  Created edge from batch_normalization/moments/SquaredDifference/squared_ to batch_normalization/moments/variance with attrs: {'in': 0, 'out': 0, 'fw_tensor_debug_info': [('batch_normalization/moments/SquaredDifference', 0)], 'in_attrs': ['in', 'control_flow_edge', 'permutation'], 'out_attrs': ['out', 'permutation'], 'data_attrs': ['fw_tensor_debug_info'], 'control_flow_edge': False}
[ 2019-08-16 11:41:46,473 ] [ DEBUG ] [ replacement:230 ]  Removing nodes: ['batch_normalization/moments/SquaredDifference']
[ 2019-08-16 11:41:46,479 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.LogSoftmax.LogSoftmaxFrontReplacer'>
[ 2019-08-16 11:41:46,498 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.LRNReplacer.LRNReplacer'>
[ 2019-08-16 11:41:46,516 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.InterpolateNormalizer.InterpolateNormalizer'>
[ 2019-08-16 11:41:46,535 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.instance_normalization.InstanceNormalization'>
[ 2019-08-16 11:41:46,553 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.image_scaler.ImageScaler'>
[ 2019-08-16 11:41:46,572 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.GNMT_DynamicSequenceLengths.GNMT_sequence_lengths'>
[ 2019-08-16 11:41:46,572 ] [ DEBUG ] [ GNMT_DynamicSequenceLengths:42 ]  +++++++++++++++ GNMT Sequence Lengths ConditionMatching ++++++++++++++++
[ 2019-08-16 11:41:46,605 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.FillToBroadcast.FillToBroadcast'>
[ 2019-08-16 11:41:46,618 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.FakeQuantWithMinMaxVars.FakeQuantWithMinMaxVarsToQuantize'>
[ INFO ]  Skip replacer <class 'extensions.front.disable_weights_quantize_value_propagation.DisableQuantizeValuePropagation'> (graph_condition not satisfied)
[ 2019-08-16 11:41:46,637 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.fake_const_ext.FakeConstToConst'>
[ 2019-08-16 11:41:46,656 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.fifo_replacer.FIFOQueue'>
[ 2019-08-16 11:41:46,675 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.override_batch.OverrideBatch'>
[ 2019-08-16 11:41:46,681 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.eltwise_n.EltwiseNReplacement'>
[ 2019-08-16 11:41:46,699 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.div.Div'>
[ 2019-08-16 11:41:46,714 ] [ DEBUG ] [ replacement:114 ]  Created edge from batch_normalization/truediv_1/mul_ to batch_normalization/cond_3/Switch_1 with attrs: {'in': 0, 'out': 0, 'fw_tensor_debug_info': [('batch_normalization/truediv_1', 0)], 'in_attrs': ['in', 'control_flow_edge', 'permutation'], 'out_attrs': ['out', 'permutation'], 'data_attrs': ['fw_tensor_debug_info'], 'control_flow_edge': False}
[ 2019-08-16 11:41:46,714 ] [ DEBUG ] [ replacement:230 ]  Removing nodes: ['batch_normalization/truediv_1']
[ 2019-08-16 11:41:46,716 ] [ DEBUG ] [ replacement:114 ]  Created edge from batch_normalization/truediv/mul_ to batch_normalization/cond_2/Switch_1 with attrs: {'in': 0, 'out': 0, 'fw_tensor_debug_info': [('batch_normalization/truediv', 0)], 'in_attrs': ['in', 'control_flow_edge', 'permutation'], 'out_attrs': ['out', 'permutation'], 'data_attrs': ['fw_tensor_debug_info'], 'control_flow_edge': False}
[ 2019-08-16 11:41:46,716 ] [ DEBUG ] [ replacement:230 ]  Removing nodes: ['batch_normalization/truediv']
[ 2019-08-16 11:41:46,722 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.CropAndResizeReplacement.CropAndResizeReplacement'>
[ 2019-08-16 11:41:46,741 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIDetectionOutputReplacement'>
[ WARNING ]  Configuration file for custom replacement with id 'ObjectDetectionAPIDetectionOutputReplacement' doesn't exist
[ INFO ]  Failed to find custom replacement description with id 'ObjectDetectionAPIDetectionOutputReplacement'
[ 2019-08-16 11:41:46,747 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.Unpack.Unpack'>
[ 2019-08-16 11:41:46,766 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.TransposeOrderNormalizer.TransposeOrderNormalizer'>
[ 2019-08-16 11:41:46,786 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIPSROIPoolingReplacement'>
[ WARNING ]  Configuration file for custom replacement with id 'ObjectDetectionAPIPSROIPoolingReplacement' doesn't exist
[ INFO ]  Failed to find custom replacement description with id 'ObjectDetectionAPIPSROIPoolingReplacement'
[ 2019-08-16 11:41:46,792 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.sub.Sub'>
[ 2019-08-16 11:41:46,807 ] [ DEBUG ] [ replacement:114 ]  Created edge from batch_normalization/sub_2/add_ to batch_normalization/truediv_1/mul_ with attrs: {'in': 1, 'out': 0, 'in_attrs': ['in', 'permutation'], 'out_attrs': ['out', 'permutation'], 'data_attrs': ['fw_tensor_debug_info']}
[ 2019-08-16 11:41:46,807 ] [ DEBUG ] [ replacement:230 ]  Removing nodes: ['batch_normalization/sub_2']
[ 2019-08-16 11:41:46,809 ] [ DEBUG ] [ replacement:114 ]  Created edge from rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/sub/add_ to rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/mul_2 with attrs: {'in': 0, 'out': 0, 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/sub', 0)], 'in_attrs': ['in', 'control_flow_edge', 'permutation'], 'out_attrs': ['out', 'permutation'], 'data_attrs': ['fw_tensor_debug_info'], 'control_flow_edge': False}
[ 2019-08-16 11:41:46,809 ] [ DEBUG ] [ replacement:230 ]  Removing nodes: ['rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/sub']
[ 2019-08-16 11:41:46,811 ] [ DEBUG ] [ replacement:114 ]  Created edge from rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/sub/add_ to rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/mul_2 with attrs: {'in': 0, 'out': 0, 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/sub', 0)], 'in_attrs': ['in', 'control_flow_edge', 'permutation'], 'out_attrs': ['out', 'permutation'], 'data_attrs': ['fw_tensor_debug_info'], 'control_flow_edge': False}
[ 2019-08-16 11:41:46,811 ] [ DEBUG ] [ replacement:230 ]  Removing nodes: ['rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/sub']
[ 2019-08-16 11:41:46,813 ] [ DEBUG ] [ replacement:114 ]  Created edge from batch_normalization/sub_1/add_ to batch_normalization/mul_1 with attrs: {'in': 0, 'out': 0, 'fw_tensor_debug_info': [('batch_normalization/sub_1', 0)], 'in_attrs': ['in', 'control_flow_edge', 'permutation'], 'out_attrs': ['out', 'permutation'], 'data_attrs': ['fw_tensor_debug_info'], 'control_flow_edge': False}
[ 2019-08-16 11:41:46,813 ] [ DEBUG ] [ replacement:230 ]  Removing nodes: ['batch_normalization/sub_1']
[ 2019-08-16 11:41:46,815 ] [ DEBUG ] [ replacement:114 ]  Created edge from batch_normalization/sub/add_ to batch_normalization/mul with attrs: {'in': 0, 'out': 0, 'fw_tensor_debug_info': [('batch_normalization/sub', 0)], 'in_attrs': ['in', 'control_flow_edge', 'permutation'], 'out_attrs': ['out', 'permutation'], 'data_attrs': ['fw_tensor_debug_info'], 'control_flow_edge': False}
[ 2019-08-16 11:41:46,815 ] [ DEBUG ] [ replacement:230 ]  Removing nodes: ['batch_normalization/sub']
[ 2019-08-16 11:41:46,817 ] [ DEBUG ] [ replacement:114 ]  Created edge from batch_normalization/batchnorm/sub/add_ to batch_normalization/batchnorm/add_1 with attrs: {'in': 1, 'out': 0, 'fw_tensor_debug_info': [('batch_normalization/batchnorm/sub', 0)], 'in_attrs': ['in', 'control_flow_edge', 'permutation'], 'out_attrs': ['out', 'permutation'], 'data_attrs': ['fw_tensor_debug_info'], 'control_flow_edge': False}
[ 2019-08-16 11:41:46,817 ] [ DEBUG ] [ replacement:230 ]  Removing nodes: ['batch_normalization/batchnorm/sub']
[ 2019-08-16 11:41:46,819 ] [ DEBUG ] [ replacement:114 ]  Created edge from rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/sub/add_ to rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/mul_2 with attrs: {'in': 0, 'out': 0, 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/sub', 0)], 'in_attrs': ['in', 'control_flow_edge', 'permutation'], 'out_attrs': ['out', 'permutation'], 'data_attrs': ['fw_tensor_debug_info'], 'control_flow_edge': False}
[ 2019-08-16 11:41:46,819 ] [ DEBUG ] [ replacement:230 ]  Removing nodes: ['rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/sub']
[ 2019-08-16 11:41:46,821 ] [ DEBUG ] [ replacement:114 ]  Created edge from rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/sub/add_ to rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/mul_2 with attrs: {'in': 0, 'out': 0, 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/sub', 0)], 'in_attrs': ['in', 'control_flow_edge', 'permutation'], 'out_attrs': ['out', 'permutation'], 'data_attrs': ['fw_tensor_debug_info'], 'control_flow_edge': False}
[ 2019-08-16 11:41:46,821 ] [ DEBUG ] [ replacement:230 ]  Removing nodes: ['rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/sub']
[ 2019-08-16 11:41:46,827 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIMaskRCNNROIPoolingSecondReplacement'>
[ WARNING ]  Configuration file for custom replacement with id 'ObjectDetectionAPIMaskRCNNROIPoolingSecondReplacement' doesn't exist
[ INFO ]  Failed to find custom replacement description with id 'ObjectDetectionAPIMaskRCNNROIPoolingSecondReplacement'
[ 2019-08-16 11:41:46,833 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIMaskRCNNSigmoidReplacement'>
[ WARNING ]  Configuration file for custom replacement with id 'ObjectDetectionAPIMaskRCNNSigmoidReplacement' doesn't exist
[ INFO ]  Failed to find custom replacement description with id 'ObjectDetectionAPIMaskRCNNSigmoidReplacement'
[ 2019-08-16 11:41:46,839 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.concat.Concat'>
[ 2019-08-16 11:41:46,858 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.CTCGreedyDecoder.CTCGreedyDecoderReplacement'>
[ 2019-08-16 11:41:46,876 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.binary_quantize_normalization.BinaryFakeQuantizeNormalization'>
[ 2019-08-16 11:41:46,928 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.basic_lstm_cell.BasicLSTMCell'>
[ 2019-08-16 11:41:46,984 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.assign_elimination.AssignSubElimination'>
[ 2019-08-16 11:41:47,003 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.assign_elimination.AssignElimination'>
[ 2019-08-16 11:41:47,023 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.assign_elimination.AssignAddElimination'>
[ 2019-08-16 11:41:47,042 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.assign_elimination.AssertElimination'>
[ 2019-08-16 11:41:47,061 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.ArgMaxSqueeze.ArgMaxSqueeze'>
[ 2019-08-16 11:41:47,081 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.tf.BlockLSTM.BlockLSTM'>
[ 2019-08-16 11:41:47,100 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.pass_separator.FrontFinish'>
[ 2019-08-16 11:41:47,106 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.front.create_tensor_nodes.CreateTensorNodes'>
[ 2019-08-16 11:41:47,148 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
seq_len/sink_port_0
[ 2019-08-16 11:41:47,253 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: batch_normalization/cond_2/ones_like
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/GRUCellZeroState/zeros
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/GRUCellZeroState/zeros
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/GRUCellZeroState/zeros
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/GRUCellZeroState/zeros
[ 2019-08-16 11:41:47,274 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.ConcatOptimization.ConcatOptimization'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,274 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.EltwiseChecker.EltwiseChecker'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,274 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.EltwiseChecker.EltwiseChecker'> will be run after <class 'extensions.middle.EltwiseInputReshape.Eltwise1DInputReshape'>
[ 2019-08-16 11:41:47,274 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.EltwiseChecker.EltwiseChecker'> will be run after <class 'extensions.middle.GemmToFullyConnected.GemmToFullyConnected'>
[ 2019-08-16 11:41:47,274 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.EltwiseInputReshape.Eltwise1DInputReshape'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,274 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.EltwiseInputReshape.Eltwise1DInputReshape'> will be run after <class 'extensions.middle.EltwiseInputReshape.EltwiseInputReshape'>
[ 2019-08-16 11:41:47,274 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.EltwiseInputReshape.EltwiseInputReshape'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,274 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.EltwiseInputReshape.EltwiseInputReshape'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,275 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.EltwiseInputNormalization.EltwiseInputNormalize'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,275 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.EltwiseInputNormalization.EltwiseInputNormalize'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,275 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.TensorIteratorCondition.LoopConditionMatcher'> will be run before <class 'extensions.middle.TensorIteratorMerge.TensorIteratorMerge'>
[ 2019-08-16 11:41:47,275 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.TensorIteratorCondition.SimpleConditionMatcher'> will be run before <class 'extensions.middle.TensorIteratorMerge.TensorIteratorMerge'>
[ 2019-08-16 11:41:47,275 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.TensorIteratorCondition.SimpleConditionMatcher'> will be run after <class 'extensions.middle.TensorIteratorCondition.LoopConditionMatcher'>
[ 2019-08-16 11:41:47,275 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.TensorIteratorCondition.DynamicDecoderConditionMatcher'> will be run before <class 'extensions.middle.TensorIteratorMerge.TensorIteratorMerge'>
[ 2019-08-16 11:41:47,275 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.TensorIteratorCondition.DynamicDecoderConditionMatcher'> will be run after <class 'extensions.middle.TensorIteratorCondition.SimpleConditionMatcher'>
[ 2019-08-16 11:41:47,275 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.AddIsCyclicAttribute.AddIsCyclicAttribute'> will be run after <class 'extensions.middle.DeleteControlFlowEdges.DeleteControlFlowEdges'>
[ 2019-08-16 11:41:47,275 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.AddMeanScaleValues.AddMeanScaleValues'> will be run before <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,275 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.AnchorToPriorBox.AnchorToPriorBoxes'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,275 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.AnchorToPriorBox.AnchorToPriorBoxes'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,275 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.CheckForCycle.CheckForCycle'> will be run after <class 'extensions.middle.TensorIteratorMerge.TensorIteratorMerge'>
[ 2019-08-16 11:41:47,275 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.DeleteNotExecutable.DeleteNotExecutable'> will be run after <class 'extensions.middle.TensorIteratorConditionChecker.ConditionChecks'>
[ 2019-08-16 11:41:47,275 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.BinarizeWeightsM1P1.BinarizeWeightsM1P1'> will be run before <class 'extensions.middle.CheckForCycle.CheckForCycle'>
[ 2019-08-16 11:41:47,275 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.BinarizeWeightsM1P1.BinarizeWeightsM1P1'> will be run before <class 'extensions.middle.DeleteNotExecutable.DeleteNotExecutable'>
[ 2019-08-16 11:41:47,275 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.BlockLSTMtoLSTMSequence.BlockLSTMtoLSTMSequence'> will be run before <class 'extensions.middle.LSTMRNNSequenceToTensorIterator.LSTMToTensorIterator'>
[ 2019-08-16 11:41:47,275 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.BlockLSTMtoLSTMSequence.BlockLSTMtoLSTMSequence'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,275 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.BlockLSTMtoLSTMSequence.BlockLSTMtoLSTMSequence'> will be run after <class 'extensions.middle.RNNSequenceNormalizeToIE.RNNSequenceNormalize'>
[ 2019-08-16 11:41:47,275 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.RemoveIdentity.RemoveIdentity'> will be run before <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,275 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.RemoveIdentity.RemoveIdentity'> will be run after <class 'extensions.middle.AddMeanScaleValues.AddMeanScaleValues'>
[ 2019-08-16 11:41:47,275 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.RemoveIdentity.RemoveDropout'> will be run before <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,275 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.RemoveIdentity.RemoveDropout'> will be run after <class 'extensions.middle.AddMeanScaleValues.AddMeanScaleValues'>
[ 2019-08-16 11:41:47,275 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.RemoveIdentity.RemoveNodesWithZeroPhase'> will be run before <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,275 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.RemoveIdentity.RemoveNodesWithZeroPhase'> will be run after <class 'extensions.middle.AddMeanScaleValues.AddMeanScaleValues'>
[ 2019-08-16 11:41:47,275 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.Cast.CastToFloatMark'> will be run before <class 'extensions.middle.RemoveIdentity.RemoveIdentity'>
[ 2019-08-16 11:41:47,275 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.Cast.CastToFloatMark'> will be run after <class 'extensions.middle.pass_separator.PreMiddleStart'>
[ 2019-08-16 11:41:47,276 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.ChangePlaceholderTypes.ChangePlaceholderTypes'> will be run before <class 'extensions.middle.ScaleInput.ScaleInput'>
[ 2019-08-16 11:41:47,276 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.ConstSwitchResolver.ConstSwitchEraser'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,276 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.ConstSwitchResolver.ConstSwitchEraser'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,276 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.ConvToBinaryConv.ConvToBinaryConv'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,276 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.ConvToBinaryConv.ConvToBinaryConv'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,276 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.ConvertGroupedStridedSlice.ConvertGroupedStridedSlice'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,276 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.ConvertGroupedStridedSlice.ConvertGroupedStridedSlice'> will be run after <class 'extensions.middle.SliceConverter.ConvertSlice'>
[ 2019-08-16 11:41:47,276 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.ConvertLayoutDependentOperations.ConvertLayoutDependentOperations'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,276 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.ConvertLayoutDependentOperations.ConvertLayoutDependentOperations'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,276 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.ConvertMultiInputConv.ConvertMultiInputConv'> will be run before <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,276 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.ConvertMultiInputConv.ConvertMultiInputConv'> will be run after <class 'extensions.middle.pass_separator.PreMiddleStart'>
[ 2019-08-16 11:41:47,276 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.CustomSubgraphCall.CustomSubgraphCall'> will be run before <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,276 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.CustomSubgraphCall.CustomSubgraphCall'> will be run after <class 'extensions.middle.pass_separator.PreMiddleStart'>
[ 2019-08-16 11:41:47,276 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.DecomposeBias.DecomposeBias'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,276 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.DecomposeBias.DecomposeBias'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,276 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.DecomposeBidirectionalRNNSequence.DecomposeBidirectionalRNNSequence'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,276 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.DecomposeBidirectionalRNNSequence.DecomposeBidirectionalRNNSequence'> will be run after <class 'extensions.middle.ONNXRNNSequenceNormalize.ONNXRNNSequenceNormalize'>
[ 2019-08-16 11:41:47,276 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.DecomposeBidirectionalRNNSequence.DecomposeBidirectionalRNNSequence'> will be run after <class 'extensions.middle.MXNetRNNSequenceNormalize.MXNetRNNSequenceNormalize'>
[ 2019-08-16 11:41:47,276 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.DeleteControlFlowEdges.DeleteControlFlowEdges'> will be run after <class 'extensions.middle.PartialInfer.PartialInfer'>
[ 2019-08-16 11:41:47,276 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.DepthToSpace.DepthToSpace'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,276 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.DepthToSpace.DepthToSpace'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,276 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.DilatedConvolution.DilatedConvolutionConverter'> will be run before <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,276 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.DilatedConvolution.DilatedConvolutionConverter'> will be run after <class 'extensions.middle.pass_separator.PreMiddleStart'>
[ 2019-08-16 11:41:47,276 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.FakeSplitOutputs.AddFakeOutputsToSplit'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,276 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.FakeSplitOutputs.AddFakeOutputsToSplit'> will be run after <class 'extensions.middle.TensorIteratorMerge.TensorIteratorMerge'>
[ 2019-08-16 11:41:47,276 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.FlattenONNXToReshape.FlattenONNXToReshape'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,276 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.FlattenONNXToReshape.FlattenONNXToReshape'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,277 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.FusedBatchNormNonConstant.FusedBatchNormNonConstant'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,277 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.FusedBatchNormNonConstant.FusedBatchNormNonConstant'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,277 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.FusedBatchNormTrainingCatch.FusedBatchNormTrainingCatch'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,277 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.FusedBatchNormTrainingCatch.FusedBatchNormTrainingCatch'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,277 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.GRURNNSequenceToTensorIterator.GRUAndRNNToTensorIterator'> will be run before <class 'extensions.middle.permute_tensor_iterator.TransposeTensorIteratorLSTM'>
[ 2019-08-16 11:41:47,277 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.GRURNNSequenceToTensorIterator.GRUAndRNNToTensorIterator'> will be run after <class 'extensions.middle.RNNSequenceNormalizeToIE.RNNSequenceNormalize'>
[ 2019-08-16 11:41:47,277 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.GatherNdNormalizer.GatherNdNormalize'> will be run before <class 'extensions.middle.BlockLSTMtoLSTMSequence.BlockLSTMtoLSTMSequence'>
[ 2019-08-16 11:41:47,277 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.GatherNdNormalizer.GatherNdNormalize'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,277 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.GemmResolver.GemmResolver'> will be run before <class 'extensions.middle.NormalizeFullyConnected.NormalizeFullyConnected'>
[ 2019-08-16 11:41:47,277 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.GemmResolver.GemmResolver'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,277 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.GemmToFullyConnected.GemmToFullyConnected'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,277 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.GemmToFullyConnected.GemmToFullyConnected'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,277 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.InputCut.MiddleInputCut'> will be run before <class 'extensions.middle.ChangePlaceholderTypes.ChangePlaceholderTypes'>
[ 2019-08-16 11:41:47,277 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.InputCut.MiddleInputCut'> will be run after <class 'extensions.middle.pass_separator.PreMiddleStart'>
[ 2019-08-16 11:41:47,277 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.L2NormToNorm.L2NormToNorm'> will be run before <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,277 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.L2NormToNorm.L2NormToNorm'> will be run after <class 'extensions.middle.pass_separator.PreMiddleStart'>
[ 2019-08-16 11:41:47,277 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.RNNSequenceNormalizeToIE.RNNSequenceNormalize'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,277 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.RNNSequenceNormalizeToIE.RNNSequenceNormalize'> will be run after <class 'extensions.middle.DecomposeBidirectionalRNNSequence.DecomposeBidirectionalRNNSequence'>
[ 2019-08-16 11:41:47,277 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.LSTMRNNSequenceToTensorIterator.LSTMToTensorIterator'> will be run before <class 'extensions.middle.permute_tensor_iterator.TransposeTensorIteratorLSTM'>
[ 2019-08-16 11:41:47,277 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.LSTMRNNSequenceToTensorIterator.LSTMToTensorIterator'> will be run after <class 'extensions.middle.RNNSequenceNormalizeToIE.RNNSequenceNormalize'>
[ 2019-08-16 11:41:47,277 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.MXNetRNNSequenceNormalize.MXNetRNNSequenceNormalize'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,277 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.MXNetRNNSequenceNormalize.MXNetRNNSequenceNormalize'> will be run after <class 'extensions.middle.MXNetSplitMultiLayers.MXNetSplitLayersToRNNSequence'>
[ 2019-08-16 11:41:47,277 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.MXNetSplitMultiLayers.MXNetSplitLayersToRNNSequence'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,277 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.MXNetSplitMultiLayers.MXNetSplitLayersToRNNSequence'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,277 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.MinimumMiddleReplacer.MinimumMiddleReplacer'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,277 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.MinimumMiddleReplacer.MinimumMiddleReplacer'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,278 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.NasNet.NasNet'> will be run after <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,278 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.NormalizeFullyConnected.NormalizeFullyConnected'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,278 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.NormalizeFullyConnected.NormalizeFullyConnected'> will be run after <class 'extensions.middle.GemmToFullyConnected.GemmToFullyConnected'>
[ 2019-08-16 11:41:47,278 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.NormalizePad.NormalizePad'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,278 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.NormalizePad.NormalizePad'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,278 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.ONNXRNNSequenceNormalize.ONNXRNNSequenceNormalize'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,278 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.ONNXRNNSequenceNormalize.ONNXRNNSequenceNormalize'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,278 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.ReluQuantizeFuse.ReluFakeQuantizeMark'> will be run before <class 'extensions.middle.SharedWeightsDuplication.SharedWeightsDuplication'>
[ 2019-08-16 11:41:47,278 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.ReluQuantizeFuse.ReluFakeQuantizeMark'> will be run after <class 'extensions.middle.BinarizeWeightsM1P1.BinarizeWeightsM1P1'>
[ 2019-08-16 11:41:47,278 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.ReluQuantizeFuse.ClampQuantizeMark'> will be run before <class 'extensions.middle.SharedWeightsDuplication.SharedWeightsDuplication'>
[ 2019-08-16 11:41:47,278 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.ReluQuantizeFuse.ClampQuantizeMark'> will be run after <class 'extensions.middle.BinarizeWeightsM1P1.BinarizeWeightsM1P1'>
[ 2019-08-16 11:41:47,278 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.ReluQuantizeFuse.ReluQuantizeFuse'> will be run before <class 'extensions.middle.SharedWeightsDuplication.SharedWeightsDuplication'>
[ 2019-08-16 11:41:47,278 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.ReluQuantizeFuse.ReluQuantizeFuse'> will be run after <class 'extensions.middle.ReluQuantizeFuse.ReluFakeQuantizeMark'>
[ 2019-08-16 11:41:47,278 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.RemoveDuplicationMemory.RemoveMemoryDuplicationPattern'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,278 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.RemoveDuplicationMemory.RemoveMemoryDuplicationPattern'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,278 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.RemoveRedundantReshapeAfterCropAndResize.RemoveRedundantReshapeAfterCropAndResize'> will be run after <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,278 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.RemoveUselessCrops.RemoveUselessCropsPattern'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,278 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.RemoveUselessCrops.RemoveUselessCropsPattern'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,278 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.ReplaceMemoryOffsetWithSplice.ReplaceMemoryOffsetNodePattern'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,278 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.ReplaceMemoryOffsetWithSplice.ReplaceMemoryOffsetNodePattern'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,278 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.ReplaceMemoryOffsetWithSplice.ReplaceMemoryOffsetWithMemoryNodePattern'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,278 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.ReplaceMemoryOffsetWithSplice.ReplaceMemoryOffsetWithMemoryNodePattern'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,278 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.ReplaceSpliceNodePattern.ReplaceSpliceNodePattern'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,278 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.ReplaceSpliceNodePattern.ReplaceSpliceNodePattern'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,278 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.ReverseTransposeNormalization.ReverseTransposeNormalization'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,278 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.ReverseTransposeNormalization.ReverseTransposeNormalization'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,278 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.ReverseV2ToReverseSequence.ReverseToReverseSequence'> will be run before <class 'extensions.middle.reverse_tensor_iterator.ReverseTensorIteratorLSTM'>
[ 2019-08-16 11:41:47,279 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.ReverseV2ToReverseSequence.ReverseToReverseSequence'> will be run after <class 'extensions.middle.PartialInfer.PartialInfer'>
[ 2019-08-16 11:41:47,279 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.ScaleInput.ScaleInput'> will be run before <class 'extensions.middle.AddMeanScaleValues.AddMeanScaleValues'>
[ 2019-08-16 11:41:47,279 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.ScaleInput.ScaleInput'> will be run after <class 'extensions.middle.pass_separator.PreMiddleStart'>
[ 2019-08-16 11:41:47,279 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.SharedWeightsDuplication.SharedWeightsDuplication'> will be run before <class 'extensions.middle.pass_separator.PreMiddleStart'>
[ 2019-08-16 11:41:47,279 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.SharedWeightsDuplication.SharedWeightsDuplication'> will be run after <class 'extensions.middle.CheckForCycle.CheckForCycle'>
[ 2019-08-16 11:41:47,279 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.ShuffleChannel.ShuffleChannel'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,279 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.ShuffleChannel.ShuffleChannel'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,279 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.SwapAxesMiddleReplacer.SwapAxisMiddleReplacer'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,279 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.SwapAxesMiddleReplacer.SwapAxisMiddleReplacer'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,279 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.TensorIteratorInput.SmartInputMatcher'> will be run before <class 'extensions.middle.TensorIteratorMerge.TensorIteratorMerge'>
[ 2019-08-16 11:41:47,279 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.TensorIteratorInput.SmartInputMatcher'> will be run after <class 'extensions.middle.AddIsCyclicAttribute.AddIsCyclicAttribute'>
[ 2019-08-16 11:41:47,279 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.TensorIteratorInput.SimpleInputMatcher'> will be run before <class 'extensions.middle.TensorIteratorMerge.TensorIteratorMerge'>
[ 2019-08-16 11:41:47,279 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.TensorIteratorInput.SimpleInputMatcher'> will be run after <class 'extensions.middle.DeleteNotExecutable.DeleteNotExecutable'>
[ 2019-08-16 11:41:47,279 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.TensorIteratorInput.BackEdgeSimpleInputMatcher'> will be run before <class 'extensions.middle.TensorIteratorMerge.TensorIteratorMerge'>
[ 2019-08-16 11:41:47,279 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.TensorIteratorInput.BackEdgeSimpleInputMatcher'> will be run after <class 'extensions.middle.TensorIteratorInput.SimpleInputMatcher'>
[ 2019-08-16 11:41:47,279 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.SwitchMergeOptimization.SwitchMergeMiddleReplacer'> will be run before <class 'extensions.middle.TensorIteratorInput.SmartInputMatcher'>
[ 2019-08-16 11:41:47,279 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.SwitchMergeOptimization.SwitchMergeMiddleReplacer'> will be run after <class 'extensions.middle.PartialInfer.PartialInfer'>
[ 2019-08-16 11:41:47,279 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.TF_lstm_cell_to_generic.TensorFlowLSTMtoGeneric'> will be run before <class 'extensions.middle.permute_tensor_iterator.TransposeTensorIteratorLSTM'>
[ 2019-08-16 11:41:47,279 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.TF_lstm_cell_to_generic.TensorFlowLSTMtoGeneric'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,279 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.TensorIteratorBackEdge.BackEdgesMatching'> will be run before <class 'extensions.middle.TensorIteratorMerge.TensorIteratorMerge'>
[ 2019-08-16 11:41:47,279 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.TensorIteratorBackEdge.BackEdgesMatching'> will be run after <class 'extensions.middle.TensorIteratorCondition.DynamicDecoderConditionMatcher'>
[ 2019-08-16 11:41:47,279 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.TensorIteratorConditionChecker.ConditionChecks'> will be run before <class 'extensions.middle.TensorIteratorMerge.TensorIteratorMerge'>
[ 2019-08-16 11:41:47,279 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.TensorIteratorConditionChecker.ConditionChecks'> will be run after <class 'extensions.middle.TensorIteratorBackEdge.BackEdgesMatching'>
[ 2019-08-16 11:41:47,279 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.TensorIteratorLSTMToLSTMSequence.TensorIteratorLSTM'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,279 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.TensorIteratorLSTMToLSTMSequence.TensorIteratorLSTM'> will be run after <class 'extensions.middle.TensorIteratorMerge.TensorIteratorMerge'>
[ 2019-08-16 11:41:47,279 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.TensorIteratorLSTMToLSTMSequence.TensorIteratorLSTM'> will be run after <class 'extensions.middle.ONNXRNNSequenceNormalize.ONNXRNNSequenceNormalize'>
[ 2019-08-16 11:41:47,279 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.TensorIteratorLSTMToLSTMSequence.TensorIteratorLSTM'> will be run after <class 'extensions.middle.TF_lstm_cell_to_generic.TensorFlowLSTMtoGeneric'>
[ 2019-08-16 11:41:47,280 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.TensorIteratorOutput.SmartOutputMatcher'> will be run before <class 'extensions.middle.TensorIteratorMerge.TensorIteratorMerge'>
[ 2019-08-16 11:41:47,280 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.TensorIteratorOutput.SmartOutputMatcher'> will be run after <class 'extensions.middle.TensorIteratorInput.SmartInputMatcher'>
[ 2019-08-16 11:41:47,280 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.TensorIteratorOutput.SimpleOutputMatcher'> will be run before <class 'extensions.middle.TensorIteratorMerge.TensorIteratorMerge'>
[ 2019-08-16 11:41:47,280 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.TensorIteratorOutput.SimpleOutputMatcher'> will be run before <class 'extensions.middle.TensorIteratorCondition.LoopConditionMatcher'>
[ 2019-08-16 11:41:47,280 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.TensorIteratorOutput.SimpleOutputMatcher'> will be run after <class 'extensions.middle.TensorIteratorOutput.SmartOutputMatcher'>
[ 2019-08-16 11:41:47,280 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.TileReplacer.TileMultipleAxisReplacer'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,280 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.TileReplacer.TileMultipleAxisReplacer'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,280 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.UpsampleToResample.UpsampleToResample'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,280 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.UpsampleToResample.UpsampleToResample'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,280 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.UselessMerge.UselessMergeEraser'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,280 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.UselessMerge.UselessMergeEraser'> will be run after <class 'extensions.middle.ConstSwitchResolver.ConstSwitchEraser'>
[ 2019-08-16 11:41:47,280 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.UselessSplitEraser.UselessSplitEraser'> will be run before <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,280 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.UselessSplitEraser.UselessSplitEraser'> will be run after <class 'extensions.middle.pass_separator.PreMiddleStart'>
[ 2019-08-16 11:41:47,280 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.UselessStridedSlice.UselessStridedSliceEraser'> will be run before <class 'extensions.middle.ConvertGroupedStridedSlice.ConvertGroupedStridedSlice'>
[ 2019-08-16 11:41:47,280 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.UselessStridedSlice.UselessStridedSliceEraser'> will be run after <class 'extensions.middle.SliceConverter.ConvertSlice'>
[ 2019-08-16 11:41:47,280 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.permute_tensor_iterator.TransposeTensorIteratorLSTM'> will be run after <class 'extensions.middle.TensorIteratorMerge.TensorIteratorMerge'>
[ 2019-08-16 11:41:47,280 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.permute_tensor_iterator.TransposeTensorIteratorLSTM'> will be run after <class 'extensions.middle.ONNXRNNSequenceNormalize.ONNXRNNSequenceNormalize'>
[ 2019-08-16 11:41:47,280 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.permute_tensor_iterator.TransposeTensorIteratorLSTM'> will be run after <class 'extensions.middle.LSTMRNNSequenceToTensorIterator.LSTMToTensorIterator'>
[ 2019-08-16 11:41:47,280 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.permute_tensor_iterator.TransposeTensorIteratorLSTM'> will be run after <class 'extensions.middle.SwapAxesMiddleReplacer.SwapAxisMiddleReplacer'>
[ 2019-08-16 11:41:47,280 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.quantize_fuses.MarkNodesToFuseUpToFakeQuantize'> will be run after <class 'extensions.middle.DeleteControlFlowEdges.DeleteControlFlowEdges'>
[ 2019-08-16 11:41:47,280 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.quantize_fuses.FakeQuantizeFuse'> will be run before <class 'extensions.middle.BinarizeWeightsM1P1.BinarizeWeightsM1P1'>
[ 2019-08-16 11:41:47,280 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.quantize_fuses.FakeQuantizeFuse'> will be run after <class 'extensions.middle.quantize_fuses.MarkNodesToFuseUpToFakeQuantize'>
[ 2019-08-16 11:41:47,280 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.reverse_tensor_iterator.ReverseTensorIteratorLSTM'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,280 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.reverse_tensor_iterator.ReverseTensorIteratorLSTM'> will be run after <class 'extensions.middle.ONNXRNNSequenceNormalize.ONNXRNNSequenceNormalize'>
[ 2019-08-16 11:41:47,280 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.reverse_tensor_iterator.ReverseTensorIteratorLSTM'> will be run after <class 'extensions.middle.permute_tensor_iterator.TransposeTensorIteratorLSTM'>
[ 2019-08-16 11:41:47,280 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.space_to_depth.SpaceToDepth'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,280 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.space_to_depth.SpaceToDepth'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,281 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.ssd_anchors_to_const.SsdAnchorsMiddleReplacer'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,281 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.ssd_anchors_to_const.SsdAnchorsMiddleReplacer'> will be run after <class 'extensions.middle.AnchorToPriorBox.AnchorToPriorBoxes'>
[ 2019-08-16 11:41:47,281 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.wights_permute_normalizer.WeightsPermuteNormalizer'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,281 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.wights_permute_normalizer.WeightsPermuteNormalizer'> will be run after <class 'extensions.middle.GemmToFullyConnected.GemmToFullyConnected'>
[ 2019-08-16 11:41:47,281 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.ArgMaxToTopK.ArgMaxToTopK'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,281 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.ArgMaxToTopK.ArgMaxToTopK'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,281 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.SliceConverter.ConvertSlice'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,281 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.SliceConverter.ConvertSlice'> will be run after <class 'extensions.middle.pass_separator.MiddleStart'>
[ 2019-08-16 11:41:47,281 ] [ DEBUG ] [ class_registration:244 ]  Replacer <class 'extensions.middle.GroupNorm.GroupNormToMVN'> will be run before <class 'extensions.middle.pass_separator.MiddleFinish'>
[ 2019-08-16 11:41:47,281 ] [ DEBUG ] [ class_registration:247 ]  Replacer <class 'extensions.middle.GroupNorm.GroupNormToMVN'> will be run after <class 'extensions.middle.EltwiseChecker.EltwiseChecker'>
[ 2019-08-16 11:41:47,285 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.PartialInfer.PartialInfer'>
[ 2019-08-16 11:41:47,309 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,309 ] [ DEBUG ] [ infer:130 ]  Partial infer for 800
[ 2019-08-16 11:41:47,309 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,309 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,309 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,309 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = -1
[ 2019-08-16 11:41:47,309 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,309 ] [ DEBUG ] [ infer:130 ]  Partial infer for 797
[ 2019-08-16 11:41:47,309 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,309 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,309 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,309 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = -1
[ 2019-08-16 11:41:47,309 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,309 ] [ DEBUG ] [ infer:130 ]  Partial infer for 794
[ 2019-08-16 11:41:47,310 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,310 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,310 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,310 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = -1
[ 2019-08-16 11:41:47,310 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,310 ] [ DEBUG ] [ infer:130 ]  Partial infer for 791
[ 2019-08-16 11:41:47,310 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,310 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,310 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,310 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = -1
[ 2019-08-16 11:41:47,310 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,310 ] [ DEBUG ] [ infer:130 ]  Partial infer for 788
[ 2019-08-16 11:41:47,310 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,310 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,310 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,310 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = -1
[ 2019-08-16 11:41:47,310 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,311 ] [ DEBUG ] [ infer:130 ]  Partial infer for 785
[ 2019-08-16 11:41:47,311 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,311 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,311 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,311 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = -1
[ 2019-08-16 11:41:47,311 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,311 ] [ DEBUG ] [ infer:130 ]  Partial infer for 782
[ 2019-08-16 11:41:47,311 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,311 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,311 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,311 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = -1
[ 2019-08-16 11:41:47,311 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,311 ] [ DEBUG ] [ infer:130 ]  Partial infer for 779
[ 2019-08-16 11:41:47,311 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,311 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,311 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,311 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = -1
[ 2019-08-16 11:41:47,311 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,312 ] [ DEBUG ] [ infer:130 ]  Partial infer for 776
[ 2019-08-16 11:41:47,312 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,312 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,312 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,312 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = -1.0
[ 2019-08-16 11:41:47,312 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,312 ] [ DEBUG ] [ infer:130 ]  Partial infer for 773
[ 2019-08-16 11:41:47,312 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,312 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,312 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,312 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = -1.0
[ 2019-08-16 11:41:47,312 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,312 ] [ DEBUG ] [ infer:130 ]  Partial infer for 766
[ 2019-08-16 11:41:47,312 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,312 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,312 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,312 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 2
[ 2019-08-16 11:41:47,312 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,313 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/moments/SquaredDifference/negate_const_
[ 2019-08-16 11:41:47,313 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,313 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,313 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,313 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = -1
[ 2019-08-16 11:41:47,313 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,313 ] [ DEBUG ] [ infer:130 ]  Partial infer for 761
[ 2019-08-16 11:41:47,313 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,313 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,313 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,313 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = -0.5
[ 2019-08-16 11:41:47,313 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,313 ] [ DEBUG ] [ infer:130 ]  Partial infer for 759
[ 2019-08-16 11:41:47,313 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,313 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,313 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,313 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0.5
[ 2019-08-16 11:41:47,313 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,314 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/moments/Squeeze_1/Dims
[ 2019-08-16 11:41:47,314 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,314 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,314 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,314 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [3], value = [0 1 2]
[ 2019-08-16 11:41:47,314 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,314 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/moments/Squeeze/Dims
[ 2019-08-16 11:41:47,314 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,314 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,314 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,314 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [3], value = [0 1 2]
[ 2019-08-16 11:41:47,314 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,314 ] [ DEBUG ] [ infer:130 ]  Partial infer for Sum/reduction_indices
[ 2019-08-16 11:41:47,314 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,314 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,315 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,315 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,315 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,315 ] [ DEBUG ] [ infer:130 ]  Partial infer for Max/reduction_indices
[ 2019-08-16 11:41:47,315 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,315 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,315 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,315 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 2
[ 2019-08-16 11:41:47,315 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,315 ] [ DEBUG ] [ infer:130 ]  Partial infer for Reshape_1/shape
[ 2019-08-16 11:41:47,315 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,315 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,315 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,315 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [3], value = [-1 28 41]
[ 2019-08-16 11:41:47,316 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,316 ] [ DEBUG ] [ infer:130 ]  Partial infer for conv2d/bias
[ 2019-08-16 11:41:47,316 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,316 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,316 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,316 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [41], value = [ 3.97828430e-01 -4.31426242e-03 -3.57889608e-02 -3.92717533e-02
 -8.06604549e-02 -8.69704634e-02...
[ 2019-08-16 11:41:47,316 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,316 ] [ DEBUG ] [ infer:130 ]  Partial infer for conv2d/bias/read
[ 2019-08-16 11:41:47,316 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,316 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,317 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [41], value = [ 3.97828430e-01 -4.31426242e-03 -3.57889608e-02 -3.92717533e-02
 -8.06604549e-02 -8.69704634e-02...
[ 2019-08-16 11:41:47,317 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,317 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [41], value = [ 3.97828430e-01 -4.31426242e-03 -3.57889608e-02 -3.92717533e-02
 -8.06604549e-02 -8.69704634e-02...
[ 2019-08-16 11:41:47,317 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,317 ] [ DEBUG ] [ infer:130 ]  Partial infer for conv2d/kernel
[ 2019-08-16 11:41:47,317 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,317 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,318 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,318 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1   1 192  41], value = [[[[ 0.10394846  0.266858   -0.28928807 ...  0.04898277 -0.1620783
    -0.08559268]
   [ 0.064426...
[ 2019-08-16 11:41:47,318 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,318 ] [ DEBUG ] [ infer:130 ]  Partial infer for conv2d/kernel/read
[ 2019-08-16 11:41:47,318 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,318 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,319 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1   1 192  41], value = [[[[ 0.10394846  0.266858   -0.28928807 ...  0.04898277 -0.1620783
    -0.08559268]
   [ 0.064426...
[ 2019-08-16 11:41:47,319 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,319 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1   1 192  41], value = [[[[ 0.10394846  0.266858   -0.28928807 ...  0.04898277 -0.1620783
    -0.08559268]
   [ 0.064426...
[ 2019-08-16 11:41:47,319 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,319 ] [ DEBUG ] [ infer:130 ]  Partial infer for Reshape/shape
[ 2019-08-16 11:41:47,319 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,319 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,319 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,320 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [4], value = [ -1  28   1 192]
[ 2019-08-16 11:41:47,320 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,320 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/concat/axis
[ 2019-08-16 11:41:47,320 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,320 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,320 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,320 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 2
[ 2019-08-16 11:41:47,320 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,320 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/ReverseV2/axis
[ 2019-08-16 11:41:47,320 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,320 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,320 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,320 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,320 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,320 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/concat_2/axis
[ 2019-08-16 11:41:47,320 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,321 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,321 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,321 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,321 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,321 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/concat_2/values_0
[ 2019-08-16 11:41:47,321 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,321 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,321 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,321 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [2], value = [1 0]
[ 2019-08-16 11:41:47,321 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,321 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/range_1/delta
[ 2019-08-16 11:41:47,321 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,321 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,321 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,321 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,321 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,321 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/range_1/start
[ 2019-08-16 11:41:47,322 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,322 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,322 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,322 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 2
[ 2019-08-16 11:41:47,322 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,322 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/Rank_1
[ 2019-08-16 11:41:47,322 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,322 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,322 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,322 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 3
[ 2019-08-16 11:41:47,322 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,322 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/range_1
[ 2019-08-16 11:41:47,322 ] [ DEBUG ] [ infer:131 ]  Op: Range
[ 2019-08-16 11:41:47,322 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,322 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 2
[ 2019-08-16 11:41:47,322 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 3
[ 2019-08-16 11:41:47,323 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [], value = 1
[ 2019-08-16 11:41:47,323 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,323 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [2]
[ 2019-08-16 11:41:47,323 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,323 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/concat_2
[ 2019-08-16 11:41:47,323 ] [ DEBUG ] [ infer:131 ]  Op: ConcatV2
[ 2019-08-16 11:41:47,323 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,323 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [2], value = [1 0]
[ 2019-08-16 11:41:47,323 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [2]
[ 2019-08-16 11:41:47,324 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,324 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [3], value = [1 0 2]
[ 2019-08-16 11:41:47,324 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,324 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArrayStack/range/delta
[ 2019-08-16 11:41:47,324 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,324 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,324 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,324 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,324 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,324 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArrayStack/range/start
[ 2019-08-16 11:41:47,324 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,324 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,324 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,324 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,324 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,324 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/gru_cell/candidate/bias
[ 2019-08-16 11:41:47,324 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,325 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,325 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,325 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [96], value = [-0.20630556 -0.2388258   0.32675105 -0.01013813  0.21144664 -0.11858168
  0.05351019  0.02979827...
[ 2019-08-16 11:41:47,325 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,325 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/gru_cell/candidate/bias/read
[ 2019-08-16 11:41:47,325 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,326 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,326 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [96], value = [-0.20630556 -0.2388258   0.32675105 -0.01013813  0.21144664 -0.11858168
  0.05351019  0.02979827...
[ 2019-08-16 11:41:47,326 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,327 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [96], value = [-0.20630556 -0.2388258   0.32675105 -0.01013813  0.21144664 -0.11858168
  0.05351019  0.02979827...
[ 2019-08-16 11:41:47,327 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,327 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/BiasAdd_1/Enter
[ 2019-08-16 11:41:47,327 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,327 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,328 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [96], value = [-0.20630556 -0.2388258   0.32675105 -0.01013813  0.21144664 -0.11858168
  0.05351019  0.02979827...
[ 2019-08-16 11:41:47,328 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,328 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [96], value = [-0.20630556 -0.2388258   0.32675105 -0.01013813  0.21144664 -0.11858168
  0.05351019  0.02979827...
[ 2019-08-16 11:41:47,329 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,329 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/gru_cell/candidate/kernel
[ 2019-08-16 11:41:47,329 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,329 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,329 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,329 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [288  96], value = [[ 0.12273635 -0.07469086  0.13450307 ... -0.38907412  0.09408648
   0.09192951]
 [-0.03125457 -0...
[ 2019-08-16 11:41:47,329 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,329 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/gru_cell/candidate/kernel/read
[ 2019-08-16 11:41:47,329 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,329 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,330 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [288  96], value = [[ 0.12273635 -0.07469086  0.13450307 ... -0.38907412  0.09408648
   0.09192951]
 [-0.03125457 -0...
[ 2019-08-16 11:41:47,330 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,330 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [288  96], value = [[ 0.12273635 -0.07469086  0.13450307 ... -0.38907412  0.09408648
   0.09192951]
 [-0.03125457 -0...
[ 2019-08-16 11:41:47,330 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,330 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/MatMul_1/Enter
[ 2019-08-16 11:41:47,330 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,331 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,331 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [288  96], value = [[ 0.12273635 -0.07469086  0.13450307 ... -0.38907412  0.09408648
   0.09192951]
 [-0.03125457 -0...
[ 2019-08-16 11:41:47,331 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,331 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [288  96], value = [[ 0.12273635 -0.07469086  0.13450307 ... -0.38907412  0.09408648
   0.09192951]
 [-0.03125457 -0...
[ 2019-08-16 11:41:47,332 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,332 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/gru_cell/gates/bias
[ 2019-08-16 11:41:47,332 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,332 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,332 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,333 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [192], value = [0.8636338  1.3107551  1.0014389  0.9294754  1.129363   0.976413
 0.6523678  0.8062803  0.8824739...
[ 2019-08-16 11:41:47,333 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,333 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/gru_cell/gates/bias/read
[ 2019-08-16 11:41:47,333 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,333 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,334 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [192], value = [0.8636338  1.3107551  1.0014389  0.9294754  1.129363   0.976413
 0.6523678  0.8062803  0.8824739...
[ 2019-08-16 11:41:47,334 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,335 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [192], value = [0.8636338  1.3107551  1.0014389  0.9294754  1.129363   0.976413
 0.6523678  0.8062803  0.8824739...
[ 2019-08-16 11:41:47,335 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,335 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/BiasAdd/Enter
[ 2019-08-16 11:41:47,335 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,336 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,337 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [192], value = [0.8636338  1.3107551  1.0014389  0.9294754  1.129363   0.976413
 0.6523678  0.8062803  0.8824739...
[ 2019-08-16 11:41:47,337 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,338 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [192], value = [0.8636338  1.3107551  1.0014389  0.9294754  1.129363   0.976413
 0.6523678  0.8062803  0.8824739...
[ 2019-08-16 11:41:47,338 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,338 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/gru_cell/gates/kernel
[ 2019-08-16 11:41:47,338 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,338 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,338 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,338 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [288 192], value = [[-1.0303181   0.29413357 -0.37989917 ...  0.24760938  0.15866868
   0.04757554]
 [-0.11078571 -0...
[ 2019-08-16 11:41:47,339 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,339 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/gru_cell/gates/kernel/read
[ 2019-08-16 11:41:47,339 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,339 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,339 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [288 192], value = [[-1.0303181   0.29413357 -0.37989917 ...  0.24760938  0.15866868
   0.04757554]
 [-0.11078571 -0...
[ 2019-08-16 11:41:47,339 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,340 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [288 192], value = [[-1.0303181   0.29413357 -0.37989917 ...  0.24760938  0.15866868
   0.04757554]
 [-0.11078571 -0...
[ 2019-08-16 11:41:47,340 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,340 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/MatMul/Enter
[ 2019-08-16 11:41:47,340 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,340 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,340 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [288 192], value = [[-1.0303181   0.29413357 -0.37989917 ...  0.24760938  0.15866868
   0.04757554]
 [-0.11078571 -0...
[ 2019-08-16 11:41:47,340 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,341 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [288 192], value = [[-1.0303181   0.29413357 -0.37989917 ...  0.24760938  0.15866868
   0.04757554]
 [-0.11078571 -0...
[ 2019-08-16 11:41:47,341 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,341 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/iteration_counter
[ 2019-08-16 11:41:47,341 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,341 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,341 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,341 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,341 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,341 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Enter
[ 2019-08-16 11:41:47,341 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,341 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,341 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,341 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,342 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,342 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,342 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Merge
[ 2019-08-16 11:41:47,342 ] [ DEBUG ] [ infer:131 ]  Op: Merge
[ 2019-08-16 11:41:47,342 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,342 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,342 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = <UNKNOWN>, value = <UNKNOWN>
[ 2019-08-16 11:41:47,342 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,342 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,342 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,342 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/Maximum/x
[ 2019-08-16 11:41:47,342 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,342 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,342 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,342 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,342 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,343 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArrayUnstack/range/delta
[ 2019-08-16 11:41:47,343 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,343 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,343 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,343 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,343 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,343 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArrayUnstack/range/start
[ 2019-08-16 11:41:47,343 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,343 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,343 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,343 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,343 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,343 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArrayUnstack/strided_slice/stack_2
[ 2019-08-16 11:41:47,343 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,343 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,343 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,343 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,344 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,344 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArrayUnstack/strided_slice/stack_1
[ 2019-08-16 11:41:47,344 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,344 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,344 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,344 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,344 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,344 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArrayUnstack/strided_slice/stack
[ 2019-08-16 11:41:47,344 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,344 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,344 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,344 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0]
[ 2019-08-16 11:41:47,344 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,344 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/time
[ 2019-08-16 11:41:47,344 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,345 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,345 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,345 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,345 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,345 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Enter_1
[ 2019-08-16 11:41:47,345 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,345 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,345 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,345 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,345 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,345 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,345 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Merge_1
[ 2019-08-16 11:41:47,345 ] [ DEBUG ] [ infer:131 ]  Op: Merge
[ 2019-08-16 11:41:47,345 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,345 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,345 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = <UNKNOWN>, value = <UNKNOWN>
[ 2019-08-16 11:41:47,345 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,346 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,346 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,346 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/strided_slice_1/stack_2
[ 2019-08-16 11:41:47,346 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,346 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,346 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,346 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,346 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,346 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/strided_slice_1/stack_1
[ 2019-08-16 11:41:47,346 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,346 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,346 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,346 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,346 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,347 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/strided_slice_1/stack
[ 2019-08-16 11:41:47,347 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,347 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,347 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,347 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0]
[ 2019-08-16 11:41:47,347 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,347 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/GRUCellZeroState/zeros/Const
[ 2019-08-16 11:41:47,347 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,347 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,347 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,347 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0.0
[ 2019-08-16 11:41:47,347 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,347 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/GRUCellZeroState/concat/axis
[ 2019-08-16 11:41:47,347 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,347 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,347 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,347 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,348 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,348 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/GRUCellZeroState/Const
[ 2019-08-16 11:41:47,348 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,348 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,348 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,348 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [96]
[ 2019-08-16 11:41:47,348 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,348 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/GRUCellZeroState/ExpandDims/dim
[ 2019-08-16 11:41:47,348 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,348 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,348 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,348 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,348 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,348 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/strided_slice/stack_2
[ 2019-08-16 11:41:47,348 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,348 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,348 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,349 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,349 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,349 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/strided_slice/stack_1
[ 2019-08-16 11:41:47,349 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,349 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,349 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,349 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [2]
[ 2019-08-16 11:41:47,349 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,349 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/strided_slice/stack
[ 2019-08-16 11:41:47,349 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,349 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,349 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,349 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,349 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,349 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/concat/axis
[ 2019-08-16 11:41:47,350 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,350 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,350 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,350 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,350 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,350 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/concat/values_0
[ 2019-08-16 11:41:47,350 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,350 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,350 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,350 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [2], value = [1 0]
[ 2019-08-16 11:41:47,350 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,350 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/range/delta
[ 2019-08-16 11:41:47,350 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,350 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,350 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,351 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,351 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,351 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/range/start
[ 2019-08-16 11:41:47,351 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,351 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,351 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,351 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 2
[ 2019-08-16 11:41:47,351 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,351 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/Rank
[ 2019-08-16 11:41:47,351 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,351 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,351 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,351 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 3
[ 2019-08-16 11:41:47,351 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,351 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/range
[ 2019-08-16 11:41:47,351 ] [ DEBUG ] [ infer:131 ]  Op: Range
[ 2019-08-16 11:41:47,352 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,352 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 2
[ 2019-08-16 11:41:47,352 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 3
[ 2019-08-16 11:41:47,352 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [], value = 1
[ 2019-08-16 11:41:47,352 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,352 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [2]
[ 2019-08-16 11:41:47,352 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,352 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/concat
[ 2019-08-16 11:41:47,352 ] [ DEBUG ] [ infer:131 ]  Op: ConcatV2
[ 2019-08-16 11:41:47,352 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,352 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [2], value = [1 0]
[ 2019-08-16 11:41:47,353 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [2]
[ 2019-08-16 11:41:47,353 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,353 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [3], value = [1 0 2]
[ 2019-08-16 11:41:47,353 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,353 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/ReverseV2/axis
[ 2019-08-16 11:41:47,353 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,353 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,353 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,353 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,353 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,353 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/concat_2/axis
[ 2019-08-16 11:41:47,353 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,353 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,353 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,353 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,354 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,354 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/concat_2/values_0
[ 2019-08-16 11:41:47,354 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,354 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,354 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,354 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [2], value = [1 0]
[ 2019-08-16 11:41:47,354 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,354 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/range_1/delta
[ 2019-08-16 11:41:47,354 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,354 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,354 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,354 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,354 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,354 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/range_1/start
[ 2019-08-16 11:41:47,354 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,354 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,354 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,355 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 2
[ 2019-08-16 11:41:47,355 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,355 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/Rank_1
[ 2019-08-16 11:41:47,355 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,355 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,355 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,355 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 3
[ 2019-08-16 11:41:47,355 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,355 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/range_1
[ 2019-08-16 11:41:47,355 ] [ DEBUG ] [ infer:131 ]  Op: Range
[ 2019-08-16 11:41:47,355 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,355 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 2
[ 2019-08-16 11:41:47,355 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 3
[ 2019-08-16 11:41:47,355 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [], value = 1
[ 2019-08-16 11:41:47,355 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,355 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [2]
[ 2019-08-16 11:41:47,356 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,356 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/concat_2
[ 2019-08-16 11:41:47,356 ] [ DEBUG ] [ infer:131 ]  Op: ConcatV2
[ 2019-08-16 11:41:47,356 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,356 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [2], value = [1 0]
[ 2019-08-16 11:41:47,356 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [2]
[ 2019-08-16 11:41:47,356 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,356 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [3], value = [1 0 2]
[ 2019-08-16 11:41:47,356 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,357 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArrayStack/range/delta
[ 2019-08-16 11:41:47,357 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,357 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,357 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,357 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,357 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,357 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArrayStack/range/start
[ 2019-08-16 11:41:47,357 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,357 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,357 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,357 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,357 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,357 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/gru_cell/candidate/bias
[ 2019-08-16 11:41:47,357 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,357 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,357 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,358 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [96], value = [ 0.14700736 -0.04902952 -0.26648712 -0.27668694  0.1791299  -0.20344824
  0.2814196  -0.05231852...
[ 2019-08-16 11:41:47,358 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,358 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/gru_cell/candidate/bias/read
[ 2019-08-16 11:41:47,358 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,358 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,359 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [96], value = [ 0.14700736 -0.04902952 -0.26648712 -0.27668694  0.1791299  -0.20344824
  0.2814196  -0.05231852...
[ 2019-08-16 11:41:47,359 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,360 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [96], value = [ 0.14700736 -0.04902952 -0.26648712 -0.27668694  0.1791299  -0.20344824
  0.2814196  -0.05231852...
[ 2019-08-16 11:41:47,360 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,360 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/BiasAdd_1/Enter
[ 2019-08-16 11:41:47,360 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,360 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,361 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [96], value = [ 0.14700736 -0.04902952 -0.26648712 -0.27668694  0.1791299  -0.20344824
  0.2814196  -0.05231852...
[ 2019-08-16 11:41:47,361 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,361 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [96], value = [ 0.14700736 -0.04902952 -0.26648712 -0.27668694  0.1791299  -0.20344824
  0.2814196  -0.05231852...
[ 2019-08-16 11:41:47,361 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,361 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/gru_cell/candidate/kernel
[ 2019-08-16 11:41:47,361 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,361 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,362 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,362 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [288  96], value = [[-0.08617257  0.00440311  0.05893075 ...  0.07117532 -0.2046704
  -0.2270496 ]
 [ 0.08344617  0....
[ 2019-08-16 11:41:47,362 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,362 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/gru_cell/candidate/kernel/read
[ 2019-08-16 11:41:47,362 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,362 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,363 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [288  96], value = [[-0.08617257  0.00440311  0.05893075 ...  0.07117532 -0.2046704
  -0.2270496 ]
 [ 0.08344617  0....
[ 2019-08-16 11:41:47,363 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,363 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [288  96], value = [[-0.08617257  0.00440311  0.05893075 ...  0.07117532 -0.2046704
  -0.2270496 ]
 [ 0.08344617  0....
[ 2019-08-16 11:41:47,363 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,363 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/MatMul_1/Enter
[ 2019-08-16 11:41:47,363 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,363 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,364 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [288  96], value = [[-0.08617257  0.00440311  0.05893075 ...  0.07117532 -0.2046704
  -0.2270496 ]
 [ 0.08344617  0....
[ 2019-08-16 11:41:47,364 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,364 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [288  96], value = [[-0.08617257  0.00440311  0.05893075 ...  0.07117532 -0.2046704
  -0.2270496 ]
 [ 0.08344617  0....
[ 2019-08-16 11:41:47,364 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,364 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/gru_cell/gates/bias
[ 2019-08-16 11:41:47,364 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,364 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,364 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,366 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [192], value = [0.8774892  0.9430615  0.661636   0.93019795 0.9606291  0.9283228
 0.99488705 0.8112845  0.851202...
[ 2019-08-16 11:41:47,366 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,366 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/gru_cell/gates/bias/read
[ 2019-08-16 11:41:47,366 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,366 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,367 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [192], value = [0.8774892  0.9430615  0.661636   0.93019795 0.9606291  0.9283228
 0.99488705 0.8112845  0.851202...
[ 2019-08-16 11:41:47,367 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,368 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [192], value = [0.8774892  0.9430615  0.661636   0.93019795 0.9606291  0.9283228
 0.99488705 0.8112845  0.851202...
[ 2019-08-16 11:41:47,368 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,368 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/BiasAdd/Enter
[ 2019-08-16 11:41:47,368 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,368 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,369 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [192], value = [0.8774892  0.9430615  0.661636   0.93019795 0.9606291  0.9283228
 0.99488705 0.8112845  0.851202...
[ 2019-08-16 11:41:47,369 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,371 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [192], value = [0.8774892  0.9430615  0.661636   0.93019795 0.9606291  0.9283228
 0.99488705 0.8112845  0.851202...
[ 2019-08-16 11:41:47,371 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,371 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/gru_cell/gates/kernel
[ 2019-08-16 11:41:47,371 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,371 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,371 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,371 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [288 192], value = [[-0.02369513 -0.41043222 -0.1847231  ...  0.17719221 -0.09955259
  -0.23327482]
 [-0.17745     0...
[ 2019-08-16 11:41:47,371 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,371 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/gru_cell/gates/kernel/read
[ 2019-08-16 11:41:47,371 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,372 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,372 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [288 192], value = [[-0.02369513 -0.41043222 -0.1847231  ...  0.17719221 -0.09955259
  -0.23327482]
 [-0.17745     0...
[ 2019-08-16 11:41:47,372 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,372 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [288 192], value = [[-0.02369513 -0.41043222 -0.1847231  ...  0.17719221 -0.09955259
  -0.23327482]
 [-0.17745     0...
[ 2019-08-16 11:41:47,372 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,373 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/MatMul/Enter
[ 2019-08-16 11:41:47,373 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,373 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,373 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [288 192], value = [[-0.02369513 -0.41043222 -0.1847231  ...  0.17719221 -0.09955259
  -0.23327482]
 [-0.17745     0...
[ 2019-08-16 11:41:47,373 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,374 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [288 192], value = [[-0.02369513 -0.41043222 -0.1847231  ...  0.17719221 -0.09955259
  -0.23327482]
 [-0.17745     0...
[ 2019-08-16 11:41:47,374 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,374 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/iteration_counter
[ 2019-08-16 11:41:47,374 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,374 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,374 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,374 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,374 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,374 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Enter
[ 2019-08-16 11:41:47,374 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,374 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,374 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,374 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,374 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,374 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,374 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Merge
[ 2019-08-16 11:41:47,375 ] [ DEBUG ] [ infer:131 ]  Op: Merge
[ 2019-08-16 11:41:47,375 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,375 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,375 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = <UNKNOWN>, value = <UNKNOWN>
[ 2019-08-16 11:41:47,375 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,375 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,375 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,375 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/Maximum/x
[ 2019-08-16 11:41:47,375 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,375 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,375 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,375 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,375 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,375 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArrayUnstack/range/delta
[ 2019-08-16 11:41:47,375 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,375 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,375 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,376 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,376 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,376 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArrayUnstack/range/start
[ 2019-08-16 11:41:47,376 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,376 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,376 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,376 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,376 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,376 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArrayUnstack/strided_slice/stack_2
[ 2019-08-16 11:41:47,376 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,376 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,376 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,376 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,376 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,376 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArrayUnstack/strided_slice/stack_1
[ 2019-08-16 11:41:47,376 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,376 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,377 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,377 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,377 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,377 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArrayUnstack/strided_slice/stack
[ 2019-08-16 11:41:47,377 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,377 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,377 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,377 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0]
[ 2019-08-16 11:41:47,377 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,377 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/time
[ 2019-08-16 11:41:47,377 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,377 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,377 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,377 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,378 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,378 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Enter_1
[ 2019-08-16 11:41:47,378 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,378 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,378 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,378 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,378 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,378 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,378 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Merge_1
[ 2019-08-16 11:41:47,378 ] [ DEBUG ] [ infer:131 ]  Op: Merge
[ 2019-08-16 11:41:47,378 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,378 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,378 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = <UNKNOWN>, value = <UNKNOWN>
[ 2019-08-16 11:41:47,378 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,378 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,378 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,379 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/strided_slice_1/stack_2
[ 2019-08-16 11:41:47,379 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,379 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,379 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,379 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,379 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,379 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/strided_slice_1/stack_1
[ 2019-08-16 11:41:47,379 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,379 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,379 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,379 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,379 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,379 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/strided_slice_1/stack
[ 2019-08-16 11:41:47,379 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,379 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,379 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,380 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0]
[ 2019-08-16 11:41:47,380 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,380 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/GRUCellZeroState/zeros/Const
[ 2019-08-16 11:41:47,380 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,380 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,380 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,380 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0.0
[ 2019-08-16 11:41:47,380 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,380 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/GRUCellZeroState/concat/axis
[ 2019-08-16 11:41:47,380 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,380 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,380 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,380 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,380 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,380 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/GRUCellZeroState/Const
[ 2019-08-16 11:41:47,380 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,381 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,381 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,381 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [96]
[ 2019-08-16 11:41:47,381 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,381 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/GRUCellZeroState/ExpandDims/dim
[ 2019-08-16 11:41:47,381 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,381 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,381 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,381 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,381 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,381 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/strided_slice/stack_2
[ 2019-08-16 11:41:47,381 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,381 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,381 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,381 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,381 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,382 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/strided_slice/stack_1
[ 2019-08-16 11:41:47,382 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,382 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,382 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,382 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [2]
[ 2019-08-16 11:41:47,382 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,382 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/strided_slice/stack
[ 2019-08-16 11:41:47,382 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,382 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,382 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,382 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,382 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,382 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/concat/axis
[ 2019-08-16 11:41:47,382 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,382 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,382 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,383 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,383 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,383 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/concat/values_0
[ 2019-08-16 11:41:47,383 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,383 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,383 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,383 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [2], value = [1 0]
[ 2019-08-16 11:41:47,383 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,383 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/range/delta
[ 2019-08-16 11:41:47,383 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,383 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,383 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,383 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,383 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,383 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/range/start
[ 2019-08-16 11:41:47,384 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,384 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,384 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,384 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 2
[ 2019-08-16 11:41:47,384 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,384 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/Rank
[ 2019-08-16 11:41:47,384 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,384 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,384 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,384 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 3
[ 2019-08-16 11:41:47,384 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,384 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/range
[ 2019-08-16 11:41:47,384 ] [ DEBUG ] [ infer:131 ]  Op: Range
[ 2019-08-16 11:41:47,384 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,384 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 2
[ 2019-08-16 11:41:47,384 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 3
[ 2019-08-16 11:41:47,385 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [], value = 1
[ 2019-08-16 11:41:47,385 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,385 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [2]
[ 2019-08-16 11:41:47,385 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,385 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/concat
[ 2019-08-16 11:41:47,385 ] [ DEBUG ] [ infer:131 ]  Op: ConcatV2
[ 2019-08-16 11:41:47,385 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,385 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [2], value = [1 0]
[ 2019-08-16 11:41:47,385 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [2]
[ 2019-08-16 11:41:47,385 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,386 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [3], value = [1 0 2]
[ 2019-08-16 11:41:47,386 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,386 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/concat/axis
[ 2019-08-16 11:41:47,386 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,386 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,386 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,386 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 2
[ 2019-08-16 11:41:47,386 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,386 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/ReverseV2/axis
[ 2019-08-16 11:41:47,386 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,386 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,386 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,386 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,386 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,386 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/concat_2/axis
[ 2019-08-16 11:41:47,386 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,386 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,387 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,387 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,387 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,387 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/concat_2/values_0
[ 2019-08-16 11:41:47,387 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,387 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,387 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,387 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [2], value = [1 0]
[ 2019-08-16 11:41:47,387 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,387 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/range_1/delta
[ 2019-08-16 11:41:47,387 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,387 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,387 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,387 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,387 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,387 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/range_1/start
[ 2019-08-16 11:41:47,387 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,388 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,388 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,388 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 2
[ 2019-08-16 11:41:47,388 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,388 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/Rank_1
[ 2019-08-16 11:41:47,388 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,388 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,388 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,388 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 3
[ 2019-08-16 11:41:47,388 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,388 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/range_1
[ 2019-08-16 11:41:47,388 ] [ DEBUG ] [ infer:131 ]  Op: Range
[ 2019-08-16 11:41:47,388 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,388 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 2
[ 2019-08-16 11:41:47,388 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 3
[ 2019-08-16 11:41:47,388 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [], value = 1
[ 2019-08-16 11:41:47,389 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,389 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [2]
[ 2019-08-16 11:41:47,389 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,389 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/concat_2
[ 2019-08-16 11:41:47,389 ] [ DEBUG ] [ infer:131 ]  Op: ConcatV2
[ 2019-08-16 11:41:47,389 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,389 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [2], value = [1 0]
[ 2019-08-16 11:41:47,389 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [2]
[ 2019-08-16 11:41:47,389 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,390 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [3], value = [1 0 2]
[ 2019-08-16 11:41:47,390 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,390 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArrayStack/range/delta
[ 2019-08-16 11:41:47,390 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,390 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,390 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,390 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,390 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,390 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArrayStack/range/start
[ 2019-08-16 11:41:47,390 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,390 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,390 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,390 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,390 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,390 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/gru_cell/candidate/bias
[ 2019-08-16 11:41:47,390 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,390 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,391 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,391 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [96], value = [ 0.14719655  0.12150294 -0.14337113 -0.34720546  0.16993123  0.01489505
  0.00510493 -0.30431303...
[ 2019-08-16 11:41:47,391 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,391 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/gru_cell/candidate/bias/read
[ 2019-08-16 11:41:47,391 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,391 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,392 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [96], value = [ 0.14719655  0.12150294 -0.14337113 -0.34720546  0.16993123  0.01489505
  0.00510493 -0.30431303...
[ 2019-08-16 11:41:47,392 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,393 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [96], value = [ 0.14719655  0.12150294 -0.14337113 -0.34720546  0.16993123  0.01489505
  0.00510493 -0.30431303...
[ 2019-08-16 11:41:47,393 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,393 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/BiasAdd_1/Enter
[ 2019-08-16 11:41:47,393 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,393 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,394 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [96], value = [ 0.14719655  0.12150294 -0.14337113 -0.34720546  0.16993123  0.01489505
  0.00510493 -0.30431303...
[ 2019-08-16 11:41:47,394 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,394 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [96], value = [ 0.14719655  0.12150294 -0.14337113 -0.34720546  0.16993123  0.01489505
  0.00510493 -0.30431303...
[ 2019-08-16 11:41:47,394 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,395 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/gru_cell/candidate/kernel
[ 2019-08-16 11:41:47,395 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,395 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,395 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,395 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [192  96], value = [[ 0.44494498  0.0179856   0.1610958  ... -0.42989033 -0.02577065
   0.42928106]
 [-0.0954172   0...
[ 2019-08-16 11:41:47,395 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,395 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/gru_cell/candidate/kernel/read
[ 2019-08-16 11:41:47,395 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,395 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,396 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [192  96], value = [[ 0.44494498  0.0179856   0.1610958  ... -0.42989033 -0.02577065
   0.42928106]
 [-0.0954172   0...
[ 2019-08-16 11:41:47,396 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,396 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [192  96], value = [[ 0.44494498  0.0179856   0.1610958  ... -0.42989033 -0.02577065
   0.42928106]
 [-0.0954172   0...
[ 2019-08-16 11:41:47,397 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,397 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/MatMul_1/Enter
[ 2019-08-16 11:41:47,397 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,397 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,397 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [192  96], value = [[ 0.44494498  0.0179856   0.1610958  ... -0.42989033 -0.02577065
   0.42928106]
 [-0.0954172   0...
[ 2019-08-16 11:41:47,397 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,398 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [192  96], value = [[ 0.44494498  0.0179856   0.1610958  ... -0.42989033 -0.02577065
   0.42928106]
 [-0.0954172   0...
[ 2019-08-16 11:41:47,398 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,398 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/gru_cell/gates/bias
[ 2019-08-16 11:41:47,398 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,398 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,398 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,399 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [192], value = [0.6209333  0.5225718  0.3840976  1.0650713  0.8585766  0.8721105
 0.8514211  0.62742203 0.872603...
[ 2019-08-16 11:41:47,399 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,399 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/gru_cell/gates/bias/read
[ 2019-08-16 11:41:47,399 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,399 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,400 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [192], value = [0.6209333  0.5225718  0.3840976  1.0650713  0.8585766  0.8721105
 0.8514211  0.62742203 0.872603...
[ 2019-08-16 11:41:47,401 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,402 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [192], value = [0.6209333  0.5225718  0.3840976  1.0650713  0.8585766  0.8721105
 0.8514211  0.62742203 0.872603...
[ 2019-08-16 11:41:47,402 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,402 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/BiasAdd/Enter
[ 2019-08-16 11:41:47,402 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,402 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,403 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [192], value = [0.6209333  0.5225718  0.3840976  1.0650713  0.8585766  0.8721105
 0.8514211  0.62742203 0.872603...
[ 2019-08-16 11:41:47,403 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,404 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [192], value = [0.6209333  0.5225718  0.3840976  1.0650713  0.8585766  0.8721105
 0.8514211  0.62742203 0.872603...
[ 2019-08-16 11:41:47,404 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,404 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/gru_cell/gates/kernel
[ 2019-08-16 11:41:47,404 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,404 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,404 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,405 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [192 192], value = [[-0.18263446  0.31629124  0.73795736 ... -0.0191342  -0.1825211
  -0.31635457]
 [-0.23331062 -0....
[ 2019-08-16 11:41:47,405 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,405 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/gru_cell/gates/kernel/read
[ 2019-08-16 11:41:47,405 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,405 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,405 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [192 192], value = [[-0.18263446  0.31629124  0.73795736 ... -0.0191342  -0.1825211
  -0.31635457]
 [-0.23331062 -0....
[ 2019-08-16 11:41:47,406 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,406 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [192 192], value = [[-0.18263446  0.31629124  0.73795736 ... -0.0191342  -0.1825211
  -0.31635457]
 [-0.23331062 -0....
[ 2019-08-16 11:41:47,406 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,406 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/MatMul/Enter
[ 2019-08-16 11:41:47,406 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,406 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,407 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [192 192], value = [[-0.18263446  0.31629124  0.73795736 ... -0.0191342  -0.1825211
  -0.31635457]
 [-0.23331062 -0....
[ 2019-08-16 11:41:47,407 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,407 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [192 192], value = [[-0.18263446  0.31629124  0.73795736 ... -0.0191342  -0.1825211
  -0.31635457]
 [-0.23331062 -0....
[ 2019-08-16 11:41:47,407 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,407 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/iteration_counter
[ 2019-08-16 11:41:47,407 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,407 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,407 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,407 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,408 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,408 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Enter
[ 2019-08-16 11:41:47,408 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,408 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,408 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,408 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,408 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,408 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,408 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Merge
[ 2019-08-16 11:41:47,408 ] [ DEBUG ] [ infer:131 ]  Op: Merge
[ 2019-08-16 11:41:47,408 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,408 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,408 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = <UNKNOWN>, value = <UNKNOWN>
[ 2019-08-16 11:41:47,408 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,408 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,408 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,408 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/Maximum/x
[ 2019-08-16 11:41:47,409 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,409 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,409 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,409 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,409 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,409 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArrayUnstack/range/delta
[ 2019-08-16 11:41:47,409 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,409 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,409 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,409 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,409 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,409 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArrayUnstack/range/start
[ 2019-08-16 11:41:47,409 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,409 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,409 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,409 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,409 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,410 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArrayUnstack/strided_slice/stack_2
[ 2019-08-16 11:41:47,410 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,410 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,410 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,410 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,410 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,410 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArrayUnstack/strided_slice/stack_1
[ 2019-08-16 11:41:47,410 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,410 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,410 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,410 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,410 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,410 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArrayUnstack/strided_slice/stack
[ 2019-08-16 11:41:47,410 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,410 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,410 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,411 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0]
[ 2019-08-16 11:41:47,411 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,411 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/time
[ 2019-08-16 11:41:47,411 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,411 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,411 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,411 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,411 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,411 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Enter_1
[ 2019-08-16 11:41:47,411 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,411 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,411 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,411 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,411 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,412 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,412 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Merge_1
[ 2019-08-16 11:41:47,412 ] [ DEBUG ] [ infer:131 ]  Op: Merge
[ 2019-08-16 11:41:47,412 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,412 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,412 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = <UNKNOWN>, value = <UNKNOWN>
[ 2019-08-16 11:41:47,412 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,412 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,412 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,412 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/strided_slice_1/stack_2
[ 2019-08-16 11:41:47,412 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,412 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,412 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,412 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,412 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,412 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/strided_slice_1/stack_1
[ 2019-08-16 11:41:47,412 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,413 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,413 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,413 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,413 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,413 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/strided_slice_1/stack
[ 2019-08-16 11:41:47,413 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,413 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,413 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,413 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0]
[ 2019-08-16 11:41:47,413 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,413 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/GRUCellZeroState/zeros/Const
[ 2019-08-16 11:41:47,413 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,413 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,413 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,413 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0.0
[ 2019-08-16 11:41:47,414 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,414 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/GRUCellZeroState/concat/axis
[ 2019-08-16 11:41:47,414 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,414 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,414 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,414 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,414 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,414 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/GRUCellZeroState/Const
[ 2019-08-16 11:41:47,414 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,414 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,414 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,414 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [96]
[ 2019-08-16 11:41:47,414 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,414 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/GRUCellZeroState/ExpandDims/dim
[ 2019-08-16 11:41:47,414 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,414 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,414 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,415 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,415 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,415 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/strided_slice/stack_2
[ 2019-08-16 11:41:47,415 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,415 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,415 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,415 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,415 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,415 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/strided_slice/stack_1
[ 2019-08-16 11:41:47,415 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,415 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,415 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,415 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [2]
[ 2019-08-16 11:41:47,415 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,415 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/strided_slice/stack
[ 2019-08-16 11:41:47,415 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,416 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,416 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,416 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,416 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,416 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/concat/axis
[ 2019-08-16 11:41:47,416 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,416 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,416 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,416 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,416 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,416 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/concat/values_0
[ 2019-08-16 11:41:47,416 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,416 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,416 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,416 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [2], value = [1 0]
[ 2019-08-16 11:41:47,417 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,417 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/range/delta
[ 2019-08-16 11:41:47,417 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,417 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,417 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,417 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,417 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,417 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/range/start
[ 2019-08-16 11:41:47,417 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,417 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,417 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,417 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 2
[ 2019-08-16 11:41:47,417 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,417 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/Rank
[ 2019-08-16 11:41:47,417 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,417 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,418 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,418 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 3
[ 2019-08-16 11:41:47,418 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,418 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/range
[ 2019-08-16 11:41:47,418 ] [ DEBUG ] [ infer:131 ]  Op: Range
[ 2019-08-16 11:41:47,418 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,418 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 2
[ 2019-08-16 11:41:47,418 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 3
[ 2019-08-16 11:41:47,418 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [], value = 1
[ 2019-08-16 11:41:47,418 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,418 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [2]
[ 2019-08-16 11:41:47,418 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,418 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/concat
[ 2019-08-16 11:41:47,418 ] [ DEBUG ] [ infer:131 ]  Op: ConcatV2
[ 2019-08-16 11:41:47,419 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,419 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [2], value = [1 0]
[ 2019-08-16 11:41:47,419 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [2]
[ 2019-08-16 11:41:47,419 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,419 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [3], value = [1 0 2]
[ 2019-08-16 11:41:47,419 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,419 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/ReverseV2/axis
[ 2019-08-16 11:41:47,419 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,419 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,419 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,419 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,420 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,420 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/concat_2/axis
[ 2019-08-16 11:41:47,420 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,420 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,420 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,420 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,420 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,420 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/concat_2/values_0
[ 2019-08-16 11:41:47,420 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,420 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,420 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,420 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [2], value = [1 0]
[ 2019-08-16 11:41:47,420 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,420 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/range_1/delta
[ 2019-08-16 11:41:47,420 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,420 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,421 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,421 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,421 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,421 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/range_1/start
[ 2019-08-16 11:41:47,421 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,421 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,421 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,421 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 2
[ 2019-08-16 11:41:47,421 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,421 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/Rank_1
[ 2019-08-16 11:41:47,421 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,421 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,421 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,421 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 3
[ 2019-08-16 11:41:47,421 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,421 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/range_1
[ 2019-08-16 11:41:47,421 ] [ DEBUG ] [ infer:131 ]  Op: Range
[ 2019-08-16 11:41:47,422 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,422 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 2
[ 2019-08-16 11:41:47,422 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 3
[ 2019-08-16 11:41:47,422 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [], value = 1
[ 2019-08-16 11:41:47,422 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,422 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [2]
[ 2019-08-16 11:41:47,422 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,422 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/concat_2
[ 2019-08-16 11:41:47,422 ] [ DEBUG ] [ infer:131 ]  Op: ConcatV2
[ 2019-08-16 11:41:47,422 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,422 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [2], value = [1 0]
[ 2019-08-16 11:41:47,423 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [2]
[ 2019-08-16 11:41:47,423 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,423 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [3], value = [1 0 2]
[ 2019-08-16 11:41:47,423 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,423 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArrayStack/range/delta
[ 2019-08-16 11:41:47,423 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,423 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,423 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,423 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,423 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,423 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArrayStack/range/start
[ 2019-08-16 11:41:47,423 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,423 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,423 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,423 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,424 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,424 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/gru_cell/candidate/bias
[ 2019-08-16 11:41:47,424 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,424 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,424 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,424 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [96], value = [ 0.22845171 -0.02988362 -0.05550991  0.35331067 -0.11666836  0.02229408
  0.20225537  0.07914946...
[ 2019-08-16 11:41:47,424 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,425 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/gru_cell/candidate/bias/read
[ 2019-08-16 11:41:47,425 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,425 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,425 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [96], value = [ 0.22845171 -0.02988362 -0.05550991  0.35331067 -0.11666836  0.02229408
  0.20225537  0.07914946...
[ 2019-08-16 11:41:47,425 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,426 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [96], value = [ 0.22845171 -0.02988362 -0.05550991  0.35331067 -0.11666836  0.02229408
  0.20225537  0.07914946...
[ 2019-08-16 11:41:47,426 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,426 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/BiasAdd_1/Enter
[ 2019-08-16 11:41:47,426 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,426 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,427 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [96], value = [ 0.22845171 -0.02988362 -0.05550991  0.35331067 -0.11666836  0.02229408
  0.20225537  0.07914946...
[ 2019-08-16 11:41:47,427 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,428 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [96], value = [ 0.22845171 -0.02988362 -0.05550991  0.35331067 -0.11666836  0.02229408
  0.20225537  0.07914946...
[ 2019-08-16 11:41:47,428 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,428 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/gru_cell/candidate/kernel
[ 2019-08-16 11:41:47,428 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,428 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,428 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,428 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [192  96], value = [[ 0.07099124  0.03735324 -0.12896395 ...  0.0217509   0.10517395
  -0.06781325]
 [ 0.56132555 -0...
[ 2019-08-16 11:41:47,428 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,428 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/gru_cell/candidate/kernel/read
[ 2019-08-16 11:41:47,428 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,429 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,429 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [192  96], value = [[ 0.07099124  0.03735324 -0.12896395 ...  0.0217509   0.10517395
  -0.06781325]
 [ 0.56132555 -0...
[ 2019-08-16 11:41:47,429 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,429 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [192  96], value = [[ 0.07099124  0.03735324 -0.12896395 ...  0.0217509   0.10517395
  -0.06781325]
 [ 0.56132555 -0...
[ 2019-08-16 11:41:47,430 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,430 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/MatMul_1/Enter
[ 2019-08-16 11:41:47,430 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,430 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,430 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [192  96], value = [[ 0.07099124  0.03735324 -0.12896395 ...  0.0217509   0.10517395
  -0.06781325]
 [ 0.56132555 -0...
[ 2019-08-16 11:41:47,430 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,431 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [192  96], value = [[ 0.07099124  0.03735324 -0.12896395 ...  0.0217509   0.10517395
  -0.06781325]
 [ 0.56132555 -0...
[ 2019-08-16 11:41:47,431 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,431 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/gru_cell/gates/bias
[ 2019-08-16 11:41:47,431 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,431 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,431 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,432 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [192], value = [0.8748921  0.7196429  0.42538187 0.73728985 0.51834875 0.23964985
 0.68675274 0.54531384 0.40408...
[ 2019-08-16 11:41:47,432 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,432 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/gru_cell/gates/bias/read
[ 2019-08-16 11:41:47,432 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,432 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,433 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [192], value = [0.8748921  0.7196429  0.42538187 0.73728985 0.51834875 0.23964985
 0.68675274 0.54531384 0.40408...
[ 2019-08-16 11:41:47,433 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,434 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [192], value = [0.8748921  0.7196429  0.42538187 0.73728985 0.51834875 0.23964985
 0.68675274 0.54531384 0.40408...
[ 2019-08-16 11:41:47,435 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,435 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/BiasAdd/Enter
[ 2019-08-16 11:41:47,435 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,435 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,436 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [192], value = [0.8748921  0.7196429  0.42538187 0.73728985 0.51834875 0.23964985
 0.68675274 0.54531384 0.40408...
[ 2019-08-16 11:41:47,436 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,437 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [192], value = [0.8748921  0.7196429  0.42538187 0.73728985 0.51834875 0.23964985
 0.68675274 0.54531384 0.40408...
[ 2019-08-16 11:41:47,437 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,437 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/gru_cell/gates/kernel
[ 2019-08-16 11:41:47,437 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,437 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,437 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,438 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [192 192], value = [[ 0.44174284 -0.10908651 -0.29394647 ... -0.13102415 -0.0380777
  -0.39950752]
 [ 0.2597118   0....
[ 2019-08-16 11:41:47,438 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,438 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/gru_cell/gates/kernel/read
[ 2019-08-16 11:41:47,438 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,438 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,438 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [192 192], value = [[ 0.44174284 -0.10908651 -0.29394647 ... -0.13102415 -0.0380777
  -0.39950752]
 [ 0.2597118   0....
[ 2019-08-16 11:41:47,438 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,439 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [192 192], value = [[ 0.44174284 -0.10908651 -0.29394647 ... -0.13102415 -0.0380777
  -0.39950752]
 [ 0.2597118   0....
[ 2019-08-16 11:41:47,439 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,439 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/MatMul/Enter
[ 2019-08-16 11:41:47,439 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,439 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,439 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [192 192], value = [[ 0.44174284 -0.10908651 -0.29394647 ... -0.13102415 -0.0380777
  -0.39950752]
 [ 0.2597118   0....
[ 2019-08-16 11:41:47,440 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,440 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [192 192], value = [[ 0.44174284 -0.10908651 -0.29394647 ... -0.13102415 -0.0380777
  -0.39950752]
 [ 0.2597118   0....
[ 2019-08-16 11:41:47,440 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,440 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/iteration_counter
[ 2019-08-16 11:41:47,440 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,440 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,440 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,440 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,440 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,440 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Enter
[ 2019-08-16 11:41:47,440 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,441 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,441 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,441 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,441 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,441 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,441 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Merge
[ 2019-08-16 11:41:47,441 ] [ DEBUG ] [ infer:131 ]  Op: Merge
[ 2019-08-16 11:41:47,441 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,441 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,441 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = <UNKNOWN>, value = <UNKNOWN>
[ 2019-08-16 11:41:47,441 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,441 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,441 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,441 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/Maximum/x
[ 2019-08-16 11:41:47,441 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,441 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,441 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,442 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,442 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,442 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArrayUnstack/range/delta
[ 2019-08-16 11:41:47,442 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,442 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,442 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,442 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,442 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,442 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArrayUnstack/range/start
[ 2019-08-16 11:41:47,442 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,442 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,442 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,442 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,442 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,442 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArrayUnstack/strided_slice/stack_2
[ 2019-08-16 11:41:47,442 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,442 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,443 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,443 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,443 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,443 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArrayUnstack/strided_slice/stack_1
[ 2019-08-16 11:41:47,443 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,443 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,443 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,443 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,443 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,443 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArrayUnstack/strided_slice/stack
[ 2019-08-16 11:41:47,443 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,443 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,443 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,444 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0]
[ 2019-08-16 11:41:47,444 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,444 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/time
[ 2019-08-16 11:41:47,444 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,444 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,444 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,444 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,444 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,444 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Enter_1
[ 2019-08-16 11:41:47,444 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,444 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,444 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,444 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,444 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,444 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,444 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Merge_1
[ 2019-08-16 11:41:47,444 ] [ DEBUG ] [ infer:131 ]  Op: Merge
[ 2019-08-16 11:41:47,445 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,445 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,445 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = <UNKNOWN>, value = <UNKNOWN>
[ 2019-08-16 11:41:47,445 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,445 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,445 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,445 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/strided_slice_1/stack_2
[ 2019-08-16 11:41:47,445 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,445 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,445 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,445 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,445 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,445 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/strided_slice_1/stack_1
[ 2019-08-16 11:41:47,445 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,445 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,445 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,446 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,446 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,446 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/strided_slice_1/stack
[ 2019-08-16 11:41:47,446 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,446 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,446 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,446 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0]
[ 2019-08-16 11:41:47,446 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,446 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/GRUCellZeroState/zeros/Const
[ 2019-08-16 11:41:47,446 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,446 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,446 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,446 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0.0
[ 2019-08-16 11:41:47,446 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,446 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/GRUCellZeroState/concat/axis
[ 2019-08-16 11:41:47,446 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,447 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,447 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,447 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,447 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,447 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/GRUCellZeroState/Const
[ 2019-08-16 11:41:47,447 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,447 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,447 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,447 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [96]
[ 2019-08-16 11:41:47,447 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,447 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/GRUCellZeroState/ExpandDims/dim
[ 2019-08-16 11:41:47,447 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,447 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,447 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,447 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,447 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,447 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/strided_slice/stack_2
[ 2019-08-16 11:41:47,448 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,448 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,448 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,448 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,448 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,448 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/strided_slice/stack_1
[ 2019-08-16 11:41:47,448 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,448 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,448 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,448 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [2]
[ 2019-08-16 11:41:47,448 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,448 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/strided_slice/stack
[ 2019-08-16 11:41:47,448 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,448 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,448 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,449 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,449 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,449 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/concat/axis
[ 2019-08-16 11:41:47,449 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,449 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,449 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,449 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,449 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,449 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/concat/values_0
[ 2019-08-16 11:41:47,449 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,449 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,449 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,449 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [2], value = [1 0]
[ 2019-08-16 11:41:47,449 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,449 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/range/delta
[ 2019-08-16 11:41:47,449 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,450 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,450 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,450 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,450 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,450 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/range/start
[ 2019-08-16 11:41:47,450 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,450 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,450 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,450 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 2
[ 2019-08-16 11:41:47,450 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,450 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/Rank
[ 2019-08-16 11:41:47,450 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,450 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,450 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,450 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 3
[ 2019-08-16 11:41:47,450 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,451 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/range
[ 2019-08-16 11:41:47,451 ] [ DEBUG ] [ infer:131 ]  Op: Range
[ 2019-08-16 11:41:47,451 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,451 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 2
[ 2019-08-16 11:41:47,451 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 3
[ 2019-08-16 11:41:47,451 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [], value = 1
[ 2019-08-16 11:41:47,451 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,451 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [2]
[ 2019-08-16 11:41:47,451 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,451 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/concat
[ 2019-08-16 11:41:47,451 ] [ DEBUG ] [ infer:131 ]  Op: ConcatV2
[ 2019-08-16 11:41:47,451 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,452 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [2], value = [1 0]
[ 2019-08-16 11:41:47,452 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [2]
[ 2019-08-16 11:41:47,452 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,452 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [3], value = [1 0 2]
[ 2019-08-16 11:41:47,452 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,452 ] [ DEBUG ] [ infer:130 ]  Partial infer for reshape_to_rnn/shape
[ 2019-08-16 11:41:47,452 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,452 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,452 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,452 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [3], value = [-1 28 96]
[ 2019-08-16 11:41:47,452 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,452 ] [ DEBUG ] [ infer:130 ]  Partial infer for pre_rnn_conv/LeakyRelu/alpha
[ 2019-08-16 11:41:47,452 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,453 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,453 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,453 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0.2
[ 2019-08-16 11:41:47,453 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,453 ] [ DEBUG ] [ infer:130 ]  Partial infer for pre_rnn_conv/bias
[ 2019-08-16 11:41:47,453 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,453 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,453 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,454 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [96], value = [-0.19918102 -0.17202342 -0.3115624   0.09023721 -0.37699068  0.01556218
 -0.45747772 -0.33676583...
[ 2019-08-16 11:41:47,454 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,454 ] [ DEBUG ] [ infer:130 ]  Partial infer for pre_rnn_conv/bias/read
[ 2019-08-16 11:41:47,454 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,454 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,454 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [96], value = [-0.19918102 -0.17202342 -0.3115624   0.09023721 -0.37699068  0.01556218
 -0.45747772 -0.33676583...
[ 2019-08-16 11:41:47,455 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,455 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [96], value = [-0.19918102 -0.17202342 -0.3115624   0.09023721 -0.37699068  0.01556218
 -0.45747772 -0.33676583...
[ 2019-08-16 11:41:47,455 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,455 ] [ DEBUG ] [ infer:130 ]  Partial infer for pre_rnn_conv/kernel
[ 2019-08-16 11:41:47,455 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,455 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,456 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,457 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1  7 32 96], value = [[[[ 6.59687936e-01  2.28611067e-01  2.13583503e-02 ...  1.51267409e-01
     9.23486054e-02 -1.10...
[ 2019-08-16 11:41:47,457 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,457 ] [ DEBUG ] [ infer:130 ]  Partial infer for pre_rnn_conv/kernel/read
[ 2019-08-16 11:41:47,457 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,457 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,459 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1  7 32 96], value = [[[[ 6.59687936e-01  2.28611067e-01  2.13583503e-02 ...  1.51267409e-01
     9.23486054e-02 -1.10...
[ 2019-08-16 11:41:47,459 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,460 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1  7 32 96], value = [[[[ 6.59687936e-01  2.28611067e-01  2.13583503e-02 ...  1.51267409e-01
     9.23486054e-02 -1.10...
[ 2019-08-16 11:41:47,460 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,460 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_2/moving_variance
[ 2019-08-16 11:41:47,460 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,460 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,461 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,461 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [32], value = [ 7.194266   1.1888463  1.080193   2.3269145  3.2647142  2.5782335
  2.3769848  1.431453   6.2463...
[ 2019-08-16 11:41:47,461 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,461 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_2/moving_variance/read
[ 2019-08-16 11:41:47,461 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,461 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,461 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [32], value = [ 7.194266   1.1888463  1.080193   2.3269145  3.2647142  2.5782335
  2.3769848  1.431453   6.2463...
[ 2019-08-16 11:41:47,462 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,462 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [32], value = [ 7.194266   1.1888463  1.080193   2.3269145  3.2647142  2.5782335
  2.3769848  1.431453   6.2463...
[ 2019-08-16 11:41:47,462 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,462 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_2/moving_mean
[ 2019-08-16 11:41:47,462 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,462 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,462 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,462 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [32], value = [ 0.9484175   0.03445841 -0.03909642  0.0317015   0.08591583  0.69247234
 -0.06837764 -0.4404067 ...
[ 2019-08-16 11:41:47,463 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,463 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_2/moving_mean/read
[ 2019-08-16 11:41:47,463 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,463 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,463 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [32], value = [ 0.9484175   0.03445841 -0.03909642  0.0317015   0.08591583  0.69247234
 -0.06837764 -0.4404067 ...
[ 2019-08-16 11:41:47,463 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,464 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [32], value = [ 0.9484175   0.03445841 -0.03909642  0.0317015   0.08591583  0.69247234
 -0.06837764 -0.4404067 ...
[ 2019-08-16 11:41:47,464 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,464 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_2/beta
[ 2019-08-16 11:41:47,464 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,464 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,464 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,464 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [32], value = [-0.01994368 -0.01121882  0.00969587  0.00824284 -0.00566241  0.02027678
 -0.00321546  0.00125897...
[ 2019-08-16 11:41:47,464 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,464 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_2/beta/read
[ 2019-08-16 11:41:47,464 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,464 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,465 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [32], value = [-0.01994368 -0.01121882  0.00969587  0.00824284 -0.00566241  0.02027678
 -0.00321546  0.00125897...
[ 2019-08-16 11:41:47,465 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,465 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [32], value = [-0.01994368 -0.01121882  0.00969587  0.00824284 -0.00566241  0.02027678
 -0.00321546  0.00125897...
[ 2019-08-16 11:41:47,465 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,465 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_2/gamma
[ 2019-08-16 11:41:47,465 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,465 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,465 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,466 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [32], value = [0.10199443 0.09186851 0.17798272 0.11306013 0.07966067 0.08508734
 0.08723275 0.09169521 0.12690...
[ 2019-08-16 11:41:47,466 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,466 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_2/gamma/read
[ 2019-08-16 11:41:47,466 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,466 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,466 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [32], value = [0.10199443 0.09186851 0.17798272 0.11306013 0.07966067 0.08508734
 0.08723275 0.09169521 0.12690...
[ 2019-08-16 11:41:47,466 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,467 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [32], value = [0.10199443 0.09186851 0.17798272 0.11306013 0.07966067 0.08508734
 0.08723275 0.09169521 0.12690...
[ 2019-08-16 11:41:47,467 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,467 ] [ DEBUG ] [ infer:130 ]  Partial infer for conv_2/LeakyRelu/alpha
[ 2019-08-16 11:41:47,467 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,467 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,467 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,467 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0.2
[ 2019-08-16 11:41:47,467 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,467 ] [ DEBUG ] [ infer:130 ]  Partial infer for conv_2/bias
[ 2019-08-16 11:41:47,467 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,467 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,467 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,468 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [32], value = [ 0.17609428  0.19918142 -0.35589856 -0.6077464  -0.39606714  0.25045902
 -0.19467422 -0.2599343 ...
[ 2019-08-16 11:41:47,468 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,468 ] [ DEBUG ] [ infer:130 ]  Partial infer for conv_2/bias/read
[ 2019-08-16 11:41:47,468 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,468 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,468 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [32], value = [ 0.17609428  0.19918142 -0.35589856 -0.6077464  -0.39606714  0.25045902
 -0.19467422 -0.2599343 ...
[ 2019-08-16 11:41:47,468 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,469 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [32], value = [ 0.17609428  0.19918142 -0.35589856 -0.6077464  -0.39606714  0.25045902
 -0.19467422 -0.2599343 ...
[ 2019-08-16 11:41:47,469 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,469 ] [ DEBUG ] [ infer:130 ]  Partial infer for conv_2/kernel
[ 2019-08-16 11:41:47,469 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,469 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,469 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,471 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 3  3 16 32], value = [[[[-0.38198754 -0.6187731   0.06409728 ...  0.01910362 -0.6564675
     0.26503855]
   [ 0.317624...
[ 2019-08-16 11:41:47,471 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,471 ] [ DEBUG ] [ infer:130 ]  Partial infer for conv_2/kernel/read
[ 2019-08-16 11:41:47,471 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,471 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,473 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 3  3 16 32], value = [[[[-0.38198754 -0.6187731   0.06409728 ...  0.01910362 -0.6564675
     0.26503855]
   [ 0.317624...
[ 2019-08-16 11:41:47,473 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,475 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 3  3 16 32], value = [[[[-0.38198754 -0.6187731   0.06409728 ...  0.01910362 -0.6564675
     0.26503855]
   [ 0.317624...
[ 2019-08-16 11:41:47,475 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,475 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_1/moving_variance
[ 2019-08-16 11:41:47,475 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,476 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,476 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,476 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [16], value = [0.37244457 0.2494403  0.16356586 0.8457516  1.0925868  0.15924428
 0.9654151  0.9373838  0.82486...
[ 2019-08-16 11:41:47,476 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,476 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_1/moving_variance/read
[ 2019-08-16 11:41:47,476 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,476 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,476 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [16], value = [0.37244457 0.2494403  0.16356586 0.8457516  1.0925868  0.15924428
 0.9654151  0.9373838  0.82486...
[ 2019-08-16 11:41:47,476 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,477 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [16], value = [0.37244457 0.2494403  0.16356586 0.8457516  1.0925868  0.15924428
 0.9654151  0.9373838  0.82486...
[ 2019-08-16 11:41:47,477 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,477 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_1/moving_mean
[ 2019-08-16 11:41:47,477 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,477 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,477 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,477 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [16], value = [0.1965222  0.16713437 0.1673043  0.42443582 0.5124033  0.29351762
 0.46435997 0.5234816  0.51540...
[ 2019-08-16 11:41:47,477 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,477 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_1/moving_mean/read
[ 2019-08-16 11:41:47,477 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,478 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,478 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [16], value = [0.1965222  0.16713437 0.1673043  0.42443582 0.5124033  0.29351762
 0.46435997 0.5234816  0.51540...
[ 2019-08-16 11:41:47,478 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,478 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [16], value = [0.1965222  0.16713437 0.1673043  0.42443582 0.5124033  0.29351762
 0.46435997 0.5234816  0.51540...
[ 2019-08-16 11:41:47,478 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,478 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_1/beta
[ 2019-08-16 11:41:47,478 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,478 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,478 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,479 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [16], value = [ 0.4807372   0.33530432  0.477206    0.2815567   0.1845824   0.4719081
 -0.12500939  0.35650358 ...
[ 2019-08-16 11:41:47,479 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,479 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_1/beta/read
[ 2019-08-16 11:41:47,479 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,479 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,479 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [16], value = [ 0.4807372   0.33530432  0.477206    0.2815567   0.1845824   0.4719081
 -0.12500939  0.35650358 ...
[ 2019-08-16 11:41:47,479 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,479 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [16], value = [ 0.4807372   0.33530432  0.477206    0.2815567   0.1845824   0.4719081
 -0.12500939  0.35650358 ...
[ 2019-08-16 11:41:47,480 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,480 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_1/gamma
[ 2019-08-16 11:41:47,480 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,480 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,480 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,480 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [16], value = [0.7362456  0.78492033 0.84287345 1.4038742  0.69683325 0.5745264
 1.5733211  1.2037237  1.222886...
[ 2019-08-16 11:41:47,480 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,480 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_1/gamma/read
[ 2019-08-16 11:41:47,480 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,480 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,481 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [16], value = [0.7362456  0.78492033 0.84287345 1.4038742  0.69683325 0.5745264
 1.5733211  1.2037237  1.222886...
[ 2019-08-16 11:41:47,481 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,481 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [16], value = [0.7362456  0.78492033 0.84287345 1.4038742  0.69683325 0.5745264
 1.5733211  1.2037237  1.222886...
[ 2019-08-16 11:41:47,481 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,481 ] [ DEBUG ] [ infer:130 ]  Partial infer for conv_1/LeakyRelu/alpha
[ 2019-08-16 11:41:47,481 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,481 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,481 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,481 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0.2
[ 2019-08-16 11:41:47,481 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,481 ] [ DEBUG ] [ infer:130 ]  Partial infer for conv_1/bias
[ 2019-08-16 11:41:47,481 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,481 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,482 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,482 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [16], value = [-0.393886   -0.33934292 -0.10791796 -0.01006103 -0.07696565  0.16393211
 -0.10199288 -0.02994025...
[ 2019-08-16 11:41:47,482 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,482 ] [ DEBUG ] [ infer:130 ]  Partial infer for conv_1/bias/read
[ 2019-08-16 11:41:47,482 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,482 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,482 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [16], value = [-0.393886   -0.33934292 -0.10791796 -0.01006103 -0.07696565  0.16393211
 -0.10199288 -0.02994025...
[ 2019-08-16 11:41:47,482 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,483 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [16], value = [-0.393886   -0.33934292 -0.10791796 -0.01006103 -0.07696565  0.16393211
 -0.10199288 -0.02994025...
[ 2019-08-16 11:41:47,483 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,483 ] [ DEBUG ] [ infer:130 ]  Partial infer for conv_1/kernel
[ 2019-08-16 11:41:47,483 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,483 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,483 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,484 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 3  3  1 16], value = [[[[-0.69081116  0.2088269   0.19459653  0.19081932 -0.12460805
    -0.20763531 -0.3709041  -0.24...
[ 2019-08-16 11:41:47,484 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,484 ] [ DEBUG ] [ infer:130 ]  Partial infer for conv_1/kernel/read
[ 2019-08-16 11:41:47,484 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,484 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,485 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 3  3  1 16], value = [[[[-0.69081116  0.2088269   0.19459653  0.19081932 -0.12460805
    -0.20763531 -0.3709041  -0.24...
[ 2019-08-16 11:41:47,485 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,486 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 3  3  1 16], value = [[[[-0.69081116  0.2088269   0.19459653  0.19081932 -0.12460805
    -0.20763531 -0.3709041  -0.24...
[ 2019-08-16 11:41:47,486 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,486 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/batchnorm/add/y
[ 2019-08-16 11:41:47,486 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,486 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,486 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,487 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0.001
[ 2019-08-16 11:41:47,487 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,487 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/sub_1/x
[ 2019-08-16 11:41:47,487 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,487 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,487 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,487 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1.0
[ 2019-08-16 11:41:47,487 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,487 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/sub/x
[ 2019-08-16 11:41:47,487 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,487 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,487 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,487 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1.0
[ 2019-08-16 11:41:47,487 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,487 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/add/y
[ 2019-08-16 11:41:47,487 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,487 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,488 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,488 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0.001
[ 2019-08-16 11:41:47,488 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,488 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/moments/variance/reduction_indices
[ 2019-08-16 11:41:47,488 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,488 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,488 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,488 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [3], value = [0 1 2]
[ 2019-08-16 11:41:47,488 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,488 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/moments/mean/reduction_indices
[ 2019-08-16 11:41:47,488 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,488 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,488 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,488 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [3], value = [0 1 2]
[ 2019-08-16 11:41:47,488 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,489 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/renorm_stddev_weight
[ 2019-08-16 11:41:47,489 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,489 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,489 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,489 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0.999997
[ 2019-08-16 11:41:47,489 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,489 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/renorm_stddev_weight/read
[ 2019-08-16 11:41:47,489 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,489 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,489 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0.999997
[ 2019-08-16 11:41:47,489 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,489 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0.999997
[ 2019-08-16 11:41:47,489 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,489 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/sub_1/negate_
[ 2019-08-16 11:41:47,489 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,489 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,490 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0.999997
[ 2019-08-16 11:41:47,490 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = -1
[ 2019-08-16 11:41:47,490 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,490 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = -0.9999970197677612
[ 2019-08-16 11:41:47,490 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,490 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/sub_1/add_
[ 2019-08-16 11:41:47,490 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,490 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,490 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = -0.9999970197677612
[ 2019-08-16 11:41:47,490 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 1.0
[ 2019-08-16 11:41:47,490 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,490 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 2.9802322387695312e-06
[ 2019-08-16 11:41:47,490 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,490 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/renorm_stddev
[ 2019-08-16 11:41:47,490 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,490 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,490 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,491 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0.29444945]
[ 2019-08-16 11:41:47,491 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,491 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/renorm_stddev/read
[ 2019-08-16 11:41:47,491 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,491 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,491 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [0.29444945]
[ 2019-08-16 11:41:47,491 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,491 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0.29444945]
[ 2019-08-16 11:41:47,491 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,491 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/renorm_mean_weight
[ 2019-08-16 11:41:47,492 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,492 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,492 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,492 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0.999997
[ 2019-08-16 11:41:47,492 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,492 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/renorm_mean_weight/read
[ 2019-08-16 11:41:47,492 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,492 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,492 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0.999997
[ 2019-08-16 11:41:47,492 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,492 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 0.999997
[ 2019-08-16 11:41:47,492 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,492 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/sub/negate_
[ 2019-08-16 11:41:47,492 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,493 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,493 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0.999997
[ 2019-08-16 11:41:47,493 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = -1
[ 2019-08-16 11:41:47,493 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,493 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = -0.9999970197677612
[ 2019-08-16 11:41:47,493 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,493 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/sub/add_
[ 2019-08-16 11:41:47,493 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,493 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,493 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = -0.9999970197677612
[ 2019-08-16 11:41:47,493 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 1.0
[ 2019-08-16 11:41:47,493 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,493 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 2.9802322387695312e-06
[ 2019-08-16 11:41:47,493 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,493 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/renorm_mean
[ 2019-08-16 11:41:47,493 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,493 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,494 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,494 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0.50882804]
[ 2019-08-16 11:41:47,494 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,494 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/renorm_mean/read
[ 2019-08-16 11:41:47,494 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,494 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,494 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [0.50882804]
[ 2019-08-16 11:41:47,494 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,494 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0.50882804]
[ 2019-08-16 11:41:47,494 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,494 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/moving_variance
[ 2019-08-16 11:41:47,495 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,495 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,495 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,495 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0.08572221]
[ 2019-08-16 11:41:47,495 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,495 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/moving_variance/read
[ 2019-08-16 11:41:47,495 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,495 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,495 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [0.08572221]
[ 2019-08-16 11:41:47,495 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,495 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0.08572221]
[ 2019-08-16 11:41:47,496 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,496 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/moving_mean
[ 2019-08-16 11:41:47,496 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,496 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,496 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,496 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0.50889]
[ 2019-08-16 11:41:47,496 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,496 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/moving_mean/read
[ 2019-08-16 11:41:47,496 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,496 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,496 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [0.50889]
[ 2019-08-16 11:41:47,496 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,497 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0.50889]
[ 2019-08-16 11:41:47,497 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,497 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/beta
[ 2019-08-16 11:41:47,497 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,497 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,497 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,497 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0.53559315]
[ 2019-08-16 11:41:47,497 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,497 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/beta/read
[ 2019-08-16 11:41:47,497 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,497 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,498 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [0.53559315]
[ 2019-08-16 11:41:47,498 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,498 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0.53559315]
[ 2019-08-16 11:41:47,498 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,498 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/gamma
[ 2019-08-16 11:41:47,498 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,498 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,498 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,498 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1.9158124]
[ 2019-08-16 11:41:47,498 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,498 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/gamma/read
[ 2019-08-16 11:41:47,498 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,499 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,499 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [1.9158124]
[ 2019-08-16 11:41:47,499 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,499 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1.9158124]
[ 2019-08-16 11:41:47,499 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,499 ] [ DEBUG ] [ infer:130 ]  Partial infer for split/split_dim
[ 2019-08-16 11:41:47,499 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,499 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,499 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,499 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 2
[ 2019-08-16 11:41:47,499 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,499 ] [ DEBUG ] [ infer:130 ]  Partial infer for inputs
[ 2019-08-16 11:41:47,499 ] [ DEBUG ] [ infer:131 ]  Op: Parameter
[ 2019-08-16 11:41:47,500 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,500 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,500 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 112  28   1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,500 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,500 ] [ DEBUG ] [ infer:130 ]  Partial infer for split
[ 2019-08-16 11:41:47,500 ] [ DEBUG ] [ infer:131 ]  Op: Split
[ 2019-08-16 11:41:47,500 ] [ DEBUG ] [ split:140 ]  input shape for split: [  1 112  28   1], should be split along 2 dim
[ 2019-08-16 11:41:47,500 ] [ DEBUG ] [ split:142 ]  split_dim_size type = <class 'numpy.int64'>
[ 2019-08-16 11:41:47,500 ] [ DEBUG ] [ split:150 ]  split_dim_size = 28, node.num_split = 1, div = 28.0, typeof div = <class 'numpy.float64'>
[ 2019-08-16 11:41:47,500 ] [ DEBUG ] [ split:118 ]  output shapes after split: [array([  1, 112,  28,   1])]
[ 2019-08-16 11:41:47,500 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,500 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [  1 112  28   1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,501 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,501 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 112  28   1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,501 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,501 ] [ DEBUG ] [ infer:130 ]  Partial infer for concat
[ 2019-08-16 11:41:47,501 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,501 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,501 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 112  28   1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,501 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,501 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 112  28   1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,501 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,501 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/moments/mean
[ 2019-08-16 11:41:47,501 ] [ DEBUG ] [ infer:131 ]  Op: ReduceMean
[ 2019-08-16 11:41:47,502 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,502 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 112  28   1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,502 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [3], value = [0 1 2]
[ 2019-08-16 11:41:47,502 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,502 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1 1 1 1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,502 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,502 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/moments/Squeeze
[ 2019-08-16 11:41:47,502 ] [ DEBUG ] [ infer:131 ]  Op: Squeeze
[ 2019-08-16 11:41:47,502 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,503 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1 1 1 1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,503 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [3], value = [0 1 2]
[ 2019-08-16 11:41:47,503 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,503 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,503 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,503 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/moments/StopGradient
[ 2019-08-16 11:41:47,503 ] [ DEBUG ] [ infer:131 ]  Op: StopGradient
[ 2019-08-16 11:41:47,503 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,503 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1 1 1 1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,503 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,503 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1 1 1 1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,503 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,503 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/moments/SquaredDifference/negate_
[ 2019-08-16 11:41:47,503 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,504 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,504 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1 1 1 1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,504 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = -1
[ 2019-08-16 11:41:47,504 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,504 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1 1 1 1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,504 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,504 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/moments/SquaredDifference/add_
[ 2019-08-16 11:41:47,504 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,504 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,504 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 112  28   1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,504 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1 1 1 1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,504 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,505 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 112  28   1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,505 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,505 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/moments/SquaredDifference/squared_
[ 2019-08-16 11:41:47,505 ] [ DEBUG ] [ infer:131 ]  Op: Pow
[ 2019-08-16 11:41:47,505 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,505 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 112  28   1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,505 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 2
[ 2019-08-16 11:41:47,505 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,505 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 112  28   1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,505 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,505 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/moments/variance
[ 2019-08-16 11:41:47,505 ] [ DEBUG ] [ infer:131 ]  Op: ReduceMean
[ 2019-08-16 11:41:47,506 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,506 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 112  28   1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,506 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [3], value = [0 1 2]
[ 2019-08-16 11:41:47,506 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,506 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1 1 1 1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,506 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,506 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/moments/Squeeze_1
[ 2019-08-16 11:41:47,506 ] [ DEBUG ] [ infer:131 ]  Op: Squeeze
[ 2019-08-16 11:41:47,507 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,507 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1 1 1 1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,507 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [3], value = [0 1 2]
[ 2019-08-16 11:41:47,507 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,507 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,507 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,507 ] [ DEBUG ] [ infer:130 ]  Partial infer for PlaceholderWithDefault/input
[ 2019-08-16 11:41:47,507 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,507 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,507 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,507 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = False
[ 2019-08-16 11:41:47,507 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,507 ] [ DEBUG ] [ infer:130 ]  Partial infer for PlaceholderWithDefault
[ 2019-08-16 11:41:47,507 ] [ DEBUG ] [ infer:131 ]  Op: PlaceholderWithDefault
[ 2019-08-16 11:41:47,508 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,508 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = False
[ 2019-08-16 11:41:47,508 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,508 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = False
[ 2019-08-16 11:41:47,508 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,508 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_2/cond/pred_id
[ 2019-08-16 11:41:47,508 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,508 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,508 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = False
[ 2019-08-16 11:41:47,508 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,508 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = False
[ 2019-08-16 11:41:47,508 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,508 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_2/cond/FusedBatchNorm_1/Switch_4
[ 2019-08-16 11:41:47,508 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,508 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,509 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [32], value = [ 7.194266   1.1888463  1.080193   2.3269145  3.2647142  2.5782335
  2.3769848  1.431453   6.2463...
[ 2019-08-16 11:41:47,509 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = False
[ 2019-08-16 11:41:47,509 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,509 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [32], value = [ 7.194266   1.1888463  1.080193   2.3269145  3.2647142  2.5782335
  2.3769848  1.431453   6.2463...
[ 2019-08-16 11:41:47,509 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,509 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_2/cond/FusedBatchNorm_1/Switch_3
[ 2019-08-16 11:41:47,509 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,510 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,510 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [32], value = [ 0.9484175   0.03445841 -0.03909642  0.0317015   0.08591583  0.69247234
 -0.06837764 -0.4404067 ...
[ 2019-08-16 11:41:47,510 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = False
[ 2019-08-16 11:41:47,510 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,510 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [32], value = [ 0.9484175   0.03445841 -0.03909642  0.0317015   0.08591583  0.69247234
 -0.06837764 -0.4404067 ...
[ 2019-08-16 11:41:47,510 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,510 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_2/cond/FusedBatchNorm_1/Switch_2
[ 2019-08-16 11:41:47,511 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,511 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,511 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [32], value = [-0.01994368 -0.01121882  0.00969587  0.00824284 -0.00566241  0.02027678
 -0.00321546  0.00125897...
[ 2019-08-16 11:41:47,511 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = False
[ 2019-08-16 11:41:47,511 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,511 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [32], value = [-0.01994368 -0.01121882  0.00969587  0.00824284 -0.00566241  0.02027678
 -0.00321546  0.00125897...
[ 2019-08-16 11:41:47,512 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,512 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_2/cond/FusedBatchNorm_1/Switch_1
[ 2019-08-16 11:41:47,512 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,512 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,512 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [32], value = [0.10199443 0.09186851 0.17798272 0.11306013 0.07966067 0.08508734
 0.08723275 0.09169521 0.12690...
[ 2019-08-16 11:41:47,512 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = False
[ 2019-08-16 11:41:47,512 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,513 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [32], value = [0.10199443 0.09186851 0.17798272 0.11306013 0.07966067 0.08508734
 0.08723275 0.09169521 0.12690...
[ 2019-08-16 11:41:47,513 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,513 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_2/cond/FusedBatchNorm/Switch_2
[ 2019-08-16 11:41:47,513 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,513 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,513 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [32], value = [-0.01994368 -0.01121882  0.00969587  0.00824284 -0.00566241  0.02027678
 -0.00321546  0.00125897...
[ 2019-08-16 11:41:47,513 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = False
[ 2019-08-16 11:41:47,513 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,514 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [32], value = [-0.01994368 -0.01121882  0.00969587  0.00824284 -0.00566241  0.02027678
 -0.00321546  0.00125897...
[ 2019-08-16 11:41:47,514 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,514 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_2/cond/FusedBatchNorm/Switch_1
[ 2019-08-16 11:41:47,514 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,514 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,514 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [32], value = [0.10199443 0.09186851 0.17798272 0.11306013 0.07966067 0.08508734
 0.08723275 0.09169521 0.12690...
[ 2019-08-16 11:41:47,514 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = False
[ 2019-08-16 11:41:47,514 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,515 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [32], value = [0.10199443 0.09186851 0.17798272 0.11306013 0.07966067 0.08508734
 0.08723275 0.09169521 0.12690...
[ 2019-08-16 11:41:47,515 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,515 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_2/cond/Switch
[ 2019-08-16 11:41:47,515 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,515 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,515 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = False
[ 2019-08-16 11:41:47,515 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = False
[ 2019-08-16 11:41:47,515 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,515 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [], value = False
[ 2019-08-16 11:41:47,515 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,516 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_2/cond/switch_t
[ 2019-08-16 11:41:47,516 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,516 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,516 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = False
[ 2019-08-16 11:41:47,516 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,516 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = False
[ 2019-08-16 11:41:47,516 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,516 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_2/cond/Const_1
[ 2019-08-16 11:41:47,516 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,516 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,516 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,516 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [0], value = []
[ 2019-08-16 11:41:47,516 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,516 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_2/cond/Const
[ 2019-08-16 11:41:47,516 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,516 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,516 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,517 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [0], value = []
[ 2019-08-16 11:41:47,517 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,517 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_1/cond/pred_id
[ 2019-08-16 11:41:47,517 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,517 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,517 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = False
[ 2019-08-16 11:41:47,517 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,517 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = False
[ 2019-08-16 11:41:47,517 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,517 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_1/cond/FusedBatchNorm_1/Switch_4
[ 2019-08-16 11:41:47,517 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,517 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,518 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [16], value = [0.37244457 0.2494403  0.16356586 0.8457516  1.0925868  0.15924428
 0.9654151  0.9373838  0.82486...
[ 2019-08-16 11:41:47,518 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = False
[ 2019-08-16 11:41:47,518 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,518 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [16], value = [0.37244457 0.2494403  0.16356586 0.8457516  1.0925868  0.15924428
 0.9654151  0.9373838  0.82486...
[ 2019-08-16 11:41:47,518 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,518 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_1/cond/FusedBatchNorm_1/Switch_3
[ 2019-08-16 11:41:47,518 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,518 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,518 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [16], value = [0.1965222  0.16713437 0.1673043  0.42443582 0.5124033  0.29351762
 0.46435997 0.5234816  0.51540...
[ 2019-08-16 11:41:47,519 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = False
[ 2019-08-16 11:41:47,519 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,519 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [16], value = [0.1965222  0.16713437 0.1673043  0.42443582 0.5124033  0.29351762
 0.46435997 0.5234816  0.51540...
[ 2019-08-16 11:41:47,519 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,519 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_1/cond/FusedBatchNorm_1/Switch_2
[ 2019-08-16 11:41:47,519 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,519 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,519 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [16], value = [ 0.4807372   0.33530432  0.477206    0.2815567   0.1845824   0.4719081
 -0.12500939  0.35650358 ...
[ 2019-08-16 11:41:47,519 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = False
[ 2019-08-16 11:41:47,520 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,520 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [16], value = [ 0.4807372   0.33530432  0.477206    0.2815567   0.1845824   0.4719081
 -0.12500939  0.35650358 ...
[ 2019-08-16 11:41:47,520 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,520 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_1/cond/FusedBatchNorm_1/Switch_1
[ 2019-08-16 11:41:47,520 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,520 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,520 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [16], value = [0.7362456  0.78492033 0.84287345 1.4038742  0.69683325 0.5745264
 1.5733211  1.2037237  1.222886...
[ 2019-08-16 11:41:47,521 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = False
[ 2019-08-16 11:41:47,521 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,521 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [16], value = [0.7362456  0.78492033 0.84287345 1.4038742  0.69683325 0.5745264
 1.5733211  1.2037237  1.222886...
[ 2019-08-16 11:41:47,521 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,521 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_1/cond/FusedBatchNorm/Switch_2
[ 2019-08-16 11:41:47,521 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,521 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,521 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [16], value = [ 0.4807372   0.33530432  0.477206    0.2815567   0.1845824   0.4719081
 -0.12500939  0.35650358 ...
[ 2019-08-16 11:41:47,521 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = False
[ 2019-08-16 11:41:47,522 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,522 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [16], value = [ 0.4807372   0.33530432  0.477206    0.2815567   0.1845824   0.4719081
 -0.12500939  0.35650358 ...
[ 2019-08-16 11:41:47,522 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,522 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_1/cond/FusedBatchNorm/Switch_1
[ 2019-08-16 11:41:47,522 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,522 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,522 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [16], value = [0.7362456  0.78492033 0.84287345 1.4038742  0.69683325 0.5745264
 1.5733211  1.2037237  1.222886...
[ 2019-08-16 11:41:47,522 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = False
[ 2019-08-16 11:41:47,522 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,523 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [16], value = [0.7362456  0.78492033 0.84287345 1.4038742  0.69683325 0.5745264
 1.5733211  1.2037237  1.222886...
[ 2019-08-16 11:41:47,523 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,523 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_1/cond/Switch
[ 2019-08-16 11:41:47,523 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,523 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,523 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = False
[ 2019-08-16 11:41:47,523 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = False
[ 2019-08-16 11:41:47,523 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,523 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [], value = False
[ 2019-08-16 11:41:47,523 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,523 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_1/cond/switch_t
[ 2019-08-16 11:41:47,523 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,524 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,524 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = False
[ 2019-08-16 11:41:47,524 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,524 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = False
[ 2019-08-16 11:41:47,524 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,524 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_1/cond/Const_1
[ 2019-08-16 11:41:47,524 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,524 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,524 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,524 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [0], value = []
[ 2019-08-16 11:41:47,524 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,524 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_1/cond/Const
[ 2019-08-16 11:41:47,524 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,524 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,524 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,524 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [0], value = []
[ 2019-08-16 11:41:47,525 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,525 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/cond_3/pred_id
[ 2019-08-16 11:41:47,525 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,525 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,525 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = False
[ 2019-08-16 11:41:47,525 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,525 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = False
[ 2019-08-16 11:41:47,525 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,525 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/cond_3/Switch
[ 2019-08-16 11:41:47,525 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,525 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,525 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = False
[ 2019-08-16 11:41:47,525 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = False
[ 2019-08-16 11:41:47,525 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,525 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = False
[ 2019-08-16 11:41:47,526 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,526 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/cond_3/switch_f
[ 2019-08-16 11:41:47,526 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,526 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,526 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = False
[ 2019-08-16 11:41:47,526 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,526 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = False
[ 2019-08-16 11:41:47,526 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,526 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/cond_3/zeros_like
[ 2019-08-16 11:41:47,526 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,526 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,526 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,526 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0.]
[ 2019-08-16 11:41:47,526 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,526 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/cond_2/pred_id
[ 2019-08-16 11:41:47,526 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,527 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,527 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = False
[ 2019-08-16 11:41:47,527 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,527 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = False
[ 2019-08-16 11:41:47,527 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,527 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/cond_2/Switch
[ 2019-08-16 11:41:47,527 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,527 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,527 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = False
[ 2019-08-16 11:41:47,527 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = False
[ 2019-08-16 11:41:47,527 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,527 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = False
[ 2019-08-16 11:41:47,527 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,527 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/cond_2/switch_f
[ 2019-08-16 11:41:47,528 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,528 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,528 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = False
[ 2019-08-16 11:41:47,528 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,528 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = False
[ 2019-08-16 11:41:47,528 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,528 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/cond_2/ones_like/Const
[ 2019-08-16 11:41:47,528 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,528 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,528 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,528 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1.0
[ 2019-08-16 11:41:47,528 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,528 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/cond_2/ones_like/Shape
[ 2019-08-16 11:41:47,528 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,528 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,528 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,529 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,529 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,529 ] [ DEBUG ] [ infer:130 ]  Partial infer for 768
[ 2019-08-16 11:41:47,529 ] [ DEBUG ] [ infer:131 ]  Op: Broadcast
[ 2019-08-16 11:41:47,529 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,529 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 1.0
[ 2019-08-16 11:41:47,529 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,529 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,529 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1.]
[ 2019-08-16 11:41:47,529 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,529 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/cond_1/pred_id
[ 2019-08-16 11:41:47,530 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,530 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,530 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = False
[ 2019-08-16 11:41:47,530 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,530 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = False
[ 2019-08-16 11:41:47,530 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,530 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/cond_1/Switch_2
[ 2019-08-16 11:41:47,530 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,530 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,530 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [0.08572221]
[ 2019-08-16 11:41:47,530 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = False
[ 2019-08-16 11:41:47,530 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,531 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0.08572221]
[ 2019-08-16 11:41:47,531 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,531 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/cond_1/Switch_1
[ 2019-08-16 11:41:47,531 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,531 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,531 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,531 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = False
[ 2019-08-16 11:41:47,531 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,531 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,531 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,531 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/cond_1/Merge
[ 2019-08-16 11:41:47,531 ] [ DEBUG ] [ infer:131 ]  Op: Merge
[ 2019-08-16 11:41:47,532 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,532 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [0.08572221]
[ 2019-08-16 11:41:47,532 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,532 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,532 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0.08572221]
[ 2019-08-16 11:41:47,532 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,532 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/batchnorm/add
[ 2019-08-16 11:41:47,532 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,532 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,533 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [0.08572221]
[ 2019-08-16 11:41:47,533 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 0.001
[ 2019-08-16 11:41:47,533 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,533 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0.08672221]
[ 2019-08-16 11:41:47,533 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,533 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/batchnorm/Rsqrt/pow_
[ 2019-08-16 11:41:47,533 ] [ DEBUG ] [ infer:131 ]  Op: Pow
[ 2019-08-16 11:41:47,533 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,533 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [0.08672221]
[ 2019-08-16 11:41:47,533 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = -0.5
[ 2019-08-16 11:41:47,533 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,534 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [3.3957431]
[ 2019-08-16 11:41:47,534 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,534 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/add
[ 2019-08-16 11:41:47,534 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,534 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,534 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [0.08572221]
[ 2019-08-16 11:41:47,534 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 0.001
[ 2019-08-16 11:41:47,534 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,534 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0.08672221]
[ 2019-08-16 11:41:47,534 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,535 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/Sqrt/pow_
[ 2019-08-16 11:41:47,535 ] [ DEBUG ] [ infer:131 ]  Op: Pow
[ 2019-08-16 11:41:47,535 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,535 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [0.08672221]
[ 2019-08-16 11:41:47,535 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 0.5
[ 2019-08-16 11:41:47,535 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,535 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0.29448634]
[ 2019-08-16 11:41:47,535 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,535 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/mul_1
[ 2019-08-16 11:41:47,535 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,535 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,535 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 2.9802322387695312e-06
[ 2019-08-16 11:41:47,536 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [0.29448634]
[ 2019-08-16 11:41:47,536 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,536 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [8.776377e-07]
[ 2019-08-16 11:41:47,536 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,536 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/add_2
[ 2019-08-16 11:41:47,536 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,536 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,536 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [0.29444945]
[ 2019-08-16 11:41:47,537 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [8.776377e-07]
[ 2019-08-16 11:41:47,537 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,537 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0.2944503]
[ 2019-08-16 11:41:47,537 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,537 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/truediv/reciprocal_
[ 2019-08-16 11:41:47,537 ] [ DEBUG ] [ infer:131 ]  Op: Pow
[ 2019-08-16 11:41:47,537 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,537 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [0.2944503]
[ 2019-08-16 11:41:47,537 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = -1.0
[ 2019-08-16 11:41:47,537 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,537 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [3.3961587]
[ 2019-08-16 11:41:47,538 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,538 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/truediv/mul_
[ 2019-08-16 11:41:47,538 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,538 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,538 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [3.3961587]
[ 2019-08-16 11:41:47,538 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [0.29448634]
[ 2019-08-16 11:41:47,538 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,538 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1.0001223]
[ 2019-08-16 11:41:47,538 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,538 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/cond_2/Switch_1
[ 2019-08-16 11:41:47,538 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,539 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,539 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [1.0001223]
[ 2019-08-16 11:41:47,539 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = False
[ 2019-08-16 11:41:47,539 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,539 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [1], value = [1.0001223]
[ 2019-08-16 11:41:47,539 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,539 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/cond_2/Merge
[ 2019-08-16 11:41:47,539 ] [ DEBUG ] [ infer:131 ]  Op: Merge
[ 2019-08-16 11:41:47,539 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,540 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [1.]
[ 2019-08-16 11:41:47,540 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [1.0001223]
[ 2019-08-16 11:41:47,540 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,540 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1.]
[ 2019-08-16 11:41:47,540 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,540 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/renorm_r
[ 2019-08-16 11:41:47,540 ] [ DEBUG ] [ infer:131 ]  Op: StopGradient
[ 2019-08-16 11:41:47,540 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,540 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [1.]
[ 2019-08-16 11:41:47,540 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,541 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,541 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,541 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/mul_2
[ 2019-08-16 11:41:47,541 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,541 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,541 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,541 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [1.9158124]
[ 2019-08-16 11:41:47,541 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,541 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,541 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,541 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/batchnorm/mul
[ 2019-08-16 11:41:47,541 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,542 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,542 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [3.3957431]
[ 2019-08-16 11:41:47,542 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,542 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,542 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,542 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,542 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/batchnorm/mul_1
[ 2019-08-16 11:41:47,542 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,542 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,542 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 112  28   1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,542 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,542 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,543 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 112  28   1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,543 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,543 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/truediv_1/reciprocal_
[ 2019-08-16 11:41:47,543 ] [ DEBUG ] [ infer:131 ]  Op: Pow
[ 2019-08-16 11:41:47,543 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,543 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [0.2944503]
[ 2019-08-16 11:41:47,543 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = -1.0
[ 2019-08-16 11:41:47,543 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,543 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [3.3961587]
[ 2019-08-16 11:41:47,543 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,543 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/cond/pred_id
[ 2019-08-16 11:41:47,544 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,544 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,544 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = False
[ 2019-08-16 11:41:47,544 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,544 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = False
[ 2019-08-16 11:41:47,544 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,544 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/cond/Switch_2
[ 2019-08-16 11:41:47,544 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,544 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,544 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [0.50889]
[ 2019-08-16 11:41:47,544 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = False
[ 2019-08-16 11:41:47,544 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,545 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0.50889]
[ 2019-08-16 11:41:47,545 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,545 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/cond/Switch_1
[ 2019-08-16 11:41:47,545 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,545 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,545 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,545 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = False
[ 2019-08-16 11:41:47,545 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,545 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,545 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,545 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/cond/Merge
[ 2019-08-16 11:41:47,545 ] [ DEBUG ] [ infer:131 ]  Op: Merge
[ 2019-08-16 11:41:47,545 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,546 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [0.50889]
[ 2019-08-16 11:41:47,546 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,546 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,546 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0.50889]
[ 2019-08-16 11:41:47,546 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,546 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/batchnorm/mul_2
[ 2019-08-16 11:41:47,546 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,546 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,546 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [0.50889]
[ 2019-08-16 11:41:47,547 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,547 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,547 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,547 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,547 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/batchnorm/sub/negate_
[ 2019-08-16 11:41:47,547 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,547 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,547 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,547 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = -1
[ 2019-08-16 11:41:47,547 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,547 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,547 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,547 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/mul
[ 2019-08-16 11:41:47,547 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,548 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,548 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 2.9802322387695312e-06
[ 2019-08-16 11:41:47,548 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [0.50889]
[ 2019-08-16 11:41:47,548 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,548 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1.5166103e-06]
[ 2019-08-16 11:41:47,548 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,548 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/add_1
[ 2019-08-16 11:41:47,548 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,548 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,548 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [0.50882804]
[ 2019-08-16 11:41:47,549 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [1.5166103e-06]
[ 2019-08-16 11:41:47,549 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,549 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0.50882953]
[ 2019-08-16 11:41:47,549 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,549 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/sub_2/negate_
[ 2019-08-16 11:41:47,549 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,549 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,549 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [0.50882953]
[ 2019-08-16 11:41:47,549 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = -1
[ 2019-08-16 11:41:47,549 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,550 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [-0.50882953]
[ 2019-08-16 11:41:47,550 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,550 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/sub_2/add_
[ 2019-08-16 11:41:47,550 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,550 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,550 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [-0.50882953]
[ 2019-08-16 11:41:47,550 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [0.50889]
[ 2019-08-16 11:41:47,550 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,550 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [6.043911e-05]
[ 2019-08-16 11:41:47,551 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,551 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/truediv_1/mul_
[ 2019-08-16 11:41:47,551 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,551 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,551 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [3.3961587]
[ 2019-08-16 11:41:47,551 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [6.043911e-05]
[ 2019-08-16 11:41:47,551 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,551 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0.00020526]
[ 2019-08-16 11:41:47,551 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,552 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/cond_3/Switch_1
[ 2019-08-16 11:41:47,552 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,552 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,552 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [0.00020526]
[ 2019-08-16 11:41:47,552 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = False
[ 2019-08-16 11:41:47,552 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,552 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [1], value = [0.00020526]
[ 2019-08-16 11:41:47,552 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,552 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/cond_3/Merge
[ 2019-08-16 11:41:47,552 ] [ DEBUG ] [ infer:131 ]  Op: Merge
[ 2019-08-16 11:41:47,552 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,553 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [0.]
[ 2019-08-16 11:41:47,553 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [0.00020526]
[ 2019-08-16 11:41:47,553 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,553 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [0.]
[ 2019-08-16 11:41:47,553 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,553 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/renorm_d
[ 2019-08-16 11:41:47,553 ] [ DEBUG ] [ infer:131 ]  Op: StopGradient
[ 2019-08-16 11:41:47,553 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,553 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [0.]
[ 2019-08-16 11:41:47,553 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,554 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,554 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,554 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/mul_3
[ 2019-08-16 11:41:47,554 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,554 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,554 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,554 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [1.9158124]
[ 2019-08-16 11:41:47,554 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,554 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,554 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,554 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/add_3
[ 2019-08-16 11:41:47,554 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,555 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,555 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,555 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [0.53559315]
[ 2019-08-16 11:41:47,555 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,555 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,555 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,555 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/batchnorm/sub/add_
[ 2019-08-16 11:41:47,555 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,555 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,555 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,555 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,555 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,555 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,556 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,556 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization/batchnorm/add_1
[ 2019-08-16 11:41:47,556 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,556 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,556 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 112  28   1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,556 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,556 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,556 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 112  28   1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,556 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,556 ] [ DEBUG ] [ infer:130 ]  Partial infer for conv_1/Conv2D
[ 2019-08-16 11:41:47,556 ] [ DEBUG ] [ infer:131 ]  Op: Conv2D
[ 2019-08-16 11:41:47,557 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,557 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 112  28   1], value = <UNKNOWN>
[ 2019-08-16 11:41:47,558 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 3  3  1 16], value = [[[[-0.69081116  0.2088269   0.19459653  0.19081932 -0.12460805
    -0.20763531 -0.3709041  -0.24...
[ 2019-08-16 11:41:47,558 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,558 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 112  28  16], value = <UNKNOWN>
[ 2019-08-16 11:41:47,558 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,558 ] [ DEBUG ] [ infer:130 ]  Partial infer for conv_1/BiasAdd
[ 2019-08-16 11:41:47,558 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,558 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,558 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 112  28  16], value = <UNKNOWN>
[ 2019-08-16 11:41:47,559 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [16], value = [-0.393886   -0.33934292 -0.10791796 -0.01006103 -0.07696565  0.16393211
 -0.10199288 -0.02994025...
[ 2019-08-16 11:41:47,559 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,559 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 112  28  16], value = <UNKNOWN>
[ 2019-08-16 11:41:47,559 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,559 ] [ DEBUG ] [ infer:130 ]  Partial infer for conv_1/LeakyRelu/mul
[ 2019-08-16 11:41:47,559 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,559 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,559 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0.2
[ 2019-08-16 11:41:47,559 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [  1 112  28  16], value = <UNKNOWN>
[ 2019-08-16 11:41:47,559 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,559 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 112  28  16], value = <UNKNOWN>
[ 2019-08-16 11:41:47,560 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,560 ] [ DEBUG ] [ infer:130 ]  Partial infer for conv_1/LeakyRelu
[ 2019-08-16 11:41:47,560 ] [ DEBUG ] [ infer:131 ]  Op: Maximum
[ 2019-08-16 11:41:47,560 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,560 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 112  28  16], value = <UNKNOWN>
[ 2019-08-16 11:41:47,560 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [  1 112  28  16], value = <UNKNOWN>
[ 2019-08-16 11:41:47,560 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,560 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 112  28  16], value = <UNKNOWN>
[ 2019-08-16 11:41:47,560 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,560 ] [ DEBUG ] [ infer:130 ]  Partial infer for max_pooling_1/AvgPool
[ 2019-08-16 11:41:47,560 ] [ DEBUG ] [ infer:131 ]  Op: AvgPool
[ 2019-08-16 11:41:47,560 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,560 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 112  28  16], value = <UNKNOWN>
[ 2019-08-16 11:41:47,561 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,561 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 56 14 16], value = <UNKNOWN>
[ 2019-08-16 11:41:47,561 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,561 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_1/cond/FusedBatchNorm_1/Switch
[ 2019-08-16 11:41:47,561 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,561 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,561 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 56 14 16], value = <UNKNOWN>
[ 2019-08-16 11:41:47,561 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = False
[ 2019-08-16 11:41:47,561 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,561 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 56 14 16], value = <UNKNOWN>
[ 2019-08-16 11:41:47,561 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,561 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_1/cond/FusedBatchNorm_1
[ 2019-08-16 11:41:47,561 ] [ DEBUG ] [ infer:131 ]  Op: FusedBatchNorm
[ 2019-08-16 11:41:47,561 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,562 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 56 14 16], value = <UNKNOWN>
[ 2019-08-16 11:41:47,562 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [16], value = [0.7362456  0.78492033 0.84287345 1.4038742  0.69683325 0.5745264
 1.5733211  1.2037237  1.222886...
[ 2019-08-16 11:41:47,562 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [16], value = [ 0.4807372   0.33530432  0.477206    0.2815567   0.1845824   0.4719081
 -0.12500939  0.35650358 ...
[ 2019-08-16 11:41:47,562 ] [ DEBUG ] [ infer:32 ]  input[3]: shape = [16], value = [0.1965222  0.16713437 0.1673043  0.42443582 0.5124033  0.29351762
 0.46435997 0.5234816  0.51540...
[ 2019-08-16 11:41:47,563 ] [ DEBUG ] [ infer:32 ]  input[4]: shape = [16], value = [0.37244457 0.2494403  0.16356586 0.8457516  1.0925868  0.15924428
 0.9654151  0.9373838  0.82486...
[ 2019-08-16 11:41:47,563 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,563 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 56 14 16], value = <UNKNOWN>
[ 2019-08-16 11:41:47,563 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,563 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_1/cond/FusedBatchNorm/Switch
[ 2019-08-16 11:41:47,563 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,563 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,563 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 56 14 16], value = <UNKNOWN>
[ 2019-08-16 11:41:47,563 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = False
[ 2019-08-16 11:41:47,563 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,563 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [ 1 56 14 16], value = <UNKNOWN>
[ 2019-08-16 11:41:47,563 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,563 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_1/cond/FusedBatchNorm
[ 2019-08-16 11:41:47,563 ] [ DEBUG ] [ infer:131 ]  Op: FusedBatchNorm
[ 2019-08-16 11:41:47,564 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,564 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 56 14 16], value = <UNKNOWN>
[ 2019-08-16 11:41:47,564 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [16], value = [0.7362456  0.78492033 0.84287345 1.4038742  0.69683325 0.5745264
 1.5733211  1.2037237  1.222886...
[ 2019-08-16 11:41:47,564 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [16], value = [ 0.4807372   0.33530432  0.477206    0.2815567   0.1845824   0.4719081
 -0.12500939  0.35650358 ...
[ 2019-08-16 11:41:47,564 ] [ DEBUG ] [ infer:32 ]  input[3]: shape = [0], value = []
[ 2019-08-16 11:41:47,564 ] [ DEBUG ] [ infer:32 ]  input[4]: shape = [0], value = []
[ 2019-08-16 11:41:47,564 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,564 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 56 14 16], value = <UNKNOWN>
[ 2019-08-16 11:41:47,565 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,565 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_1/cond/Merge
[ 2019-08-16 11:41:47,565 ] [ DEBUG ] [ infer:131 ]  Op: Merge
[ 2019-08-16 11:41:47,565 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,565 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 56 14 16], value = <UNKNOWN>
[ 2019-08-16 11:41:47,565 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 56 14 16], value = <UNKNOWN>
[ 2019-08-16 11:41:47,565 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,565 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 56 14 16], value = <UNKNOWN>
[ 2019-08-16 11:41:47,565 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,565 ] [ DEBUG ] [ infer:130 ]  Partial infer for conv_2/Conv2D
[ 2019-08-16 11:41:47,565 ] [ DEBUG ] [ infer:131 ]  Op: Conv2D
[ 2019-08-16 11:41:47,566 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,566 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 56 14 16], value = <UNKNOWN>
[ 2019-08-16 11:41:47,568 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 3  3 16 32], value = [[[[-0.38198754 -0.6187731   0.06409728 ...  0.01910362 -0.6564675
     0.26503855]
   [ 0.317624...
[ 2019-08-16 11:41:47,568 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,568 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 56 14 32], value = <UNKNOWN>
[ 2019-08-16 11:41:47,568 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,568 ] [ DEBUG ] [ infer:130 ]  Partial infer for conv_2/BiasAdd
[ 2019-08-16 11:41:47,568 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,568 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,569 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 56 14 32], value = <UNKNOWN>
[ 2019-08-16 11:41:47,569 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [32], value = [ 0.17609428  0.19918142 -0.35589856 -0.6077464  -0.39606714  0.25045902
 -0.19467422 -0.2599343 ...
[ 2019-08-16 11:41:47,569 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,569 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 56 14 32], value = <UNKNOWN>
[ 2019-08-16 11:41:47,569 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,569 ] [ DEBUG ] [ infer:130 ]  Partial infer for conv_2/LeakyRelu/mul
[ 2019-08-16 11:41:47,569 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,569 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,569 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0.2
[ 2019-08-16 11:41:47,569 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 56 14 32], value = <UNKNOWN>
[ 2019-08-16 11:41:47,569 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,570 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 56 14 32], value = <UNKNOWN>
[ 2019-08-16 11:41:47,570 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,570 ] [ DEBUG ] [ infer:130 ]  Partial infer for conv_2/LeakyRelu
[ 2019-08-16 11:41:47,570 ] [ DEBUG ] [ infer:131 ]  Op: Maximum
[ 2019-08-16 11:41:47,570 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,570 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 56 14 32], value = <UNKNOWN>
[ 2019-08-16 11:41:47,570 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 56 14 32], value = <UNKNOWN>
[ 2019-08-16 11:41:47,570 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,570 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 56 14 32], value = <UNKNOWN>
[ 2019-08-16 11:41:47,570 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,570 ] [ DEBUG ] [ infer:130 ]  Partial infer for max_pooling_2/AvgPool
[ 2019-08-16 11:41:47,570 ] [ DEBUG ] [ infer:131 ]  Op: AvgPool
[ 2019-08-16 11:41:47,571 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,571 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 56 14 32], value = <UNKNOWN>
[ 2019-08-16 11:41:47,571 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,571 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 28  7 32], value = <UNKNOWN>
[ 2019-08-16 11:41:47,571 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,571 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_2/cond/FusedBatchNorm_1/Switch
[ 2019-08-16 11:41:47,571 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,571 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,571 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 28  7 32], value = <UNKNOWN>
[ 2019-08-16 11:41:47,571 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = False
[ 2019-08-16 11:41:47,571 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,571 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 28  7 32], value = <UNKNOWN>
[ 2019-08-16 11:41:47,571 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,572 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_2/cond/FusedBatchNorm_1
[ 2019-08-16 11:41:47,572 ] [ DEBUG ] [ infer:131 ]  Op: FusedBatchNorm
[ 2019-08-16 11:41:47,572 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,572 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 28  7 32], value = <UNKNOWN>
[ 2019-08-16 11:41:47,572 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [32], value = [0.10199443 0.09186851 0.17798272 0.11306013 0.07966067 0.08508734
 0.08723275 0.09169521 0.12690...
[ 2019-08-16 11:41:47,572 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [32], value = [-0.01994368 -0.01121882  0.00969587  0.00824284 -0.00566241  0.02027678
 -0.00321546  0.00125897...
[ 2019-08-16 11:41:47,573 ] [ DEBUG ] [ infer:32 ]  input[3]: shape = [32], value = [ 0.9484175   0.03445841 -0.03909642  0.0317015   0.08591583  0.69247234
 -0.06837764 -0.4404067 ...
[ 2019-08-16 11:41:47,573 ] [ DEBUG ] [ infer:32 ]  input[4]: shape = [32], value = [ 7.194266   1.1888463  1.080193   2.3269145  3.2647142  2.5782335
  2.3769848  1.431453   6.2463...
[ 2019-08-16 11:41:47,573 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,573 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 28  7 32], value = <UNKNOWN>
[ 2019-08-16 11:41:47,573 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,573 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_2/cond/FusedBatchNorm/Switch
[ 2019-08-16 11:41:47,573 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,573 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,574 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 28  7 32], value = <UNKNOWN>
[ 2019-08-16 11:41:47,574 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = False
[ 2019-08-16 11:41:47,574 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,574 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [ 1 28  7 32], value = <UNKNOWN>
[ 2019-08-16 11:41:47,574 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,574 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_2/cond/FusedBatchNorm
[ 2019-08-16 11:41:47,574 ] [ DEBUG ] [ infer:131 ]  Op: FusedBatchNorm
[ 2019-08-16 11:41:47,574 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,574 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 28  7 32], value = <UNKNOWN>
[ 2019-08-16 11:41:47,574 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [32], value = [0.10199443 0.09186851 0.17798272 0.11306013 0.07966067 0.08508734
 0.08723275 0.09169521 0.12690...
[ 2019-08-16 11:41:47,575 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [32], value = [-0.01994368 -0.01121882  0.00969587  0.00824284 -0.00566241  0.02027678
 -0.00321546  0.00125897...
[ 2019-08-16 11:41:47,575 ] [ DEBUG ] [ infer:32 ]  input[3]: shape = [0], value = []
[ 2019-08-16 11:41:47,575 ] [ DEBUG ] [ infer:32 ]  input[4]: shape = [0], value = []
[ 2019-08-16 11:41:47,575 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,575 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 28  7 32], value = <UNKNOWN>
[ 2019-08-16 11:41:47,575 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,575 ] [ DEBUG ] [ infer:130 ]  Partial infer for batch_normalization_2/cond/Merge
[ 2019-08-16 11:41:47,575 ] [ DEBUG ] [ infer:131 ]  Op: Merge
[ 2019-08-16 11:41:47,575 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,576 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 28  7 32], value = <UNKNOWN>
[ 2019-08-16 11:41:47,576 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 28  7 32], value = <UNKNOWN>
[ 2019-08-16 11:41:47,576 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,576 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 28  7 32], value = <UNKNOWN>
[ 2019-08-16 11:41:47,576 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,576 ] [ DEBUG ] [ infer:130 ]  Partial infer for pre_rnn_conv/Conv2D
[ 2019-08-16 11:41:47,576 ] [ DEBUG ] [ infer:131 ]  Op: Conv2D
[ 2019-08-16 11:41:47,576 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,577 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 28  7 32], value = <UNKNOWN>
[ 2019-08-16 11:41:47,578 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1  7 32 96], value = [[[[ 6.59687936e-01  2.28611067e-01  2.13583503e-02 ...  1.51267409e-01
     9.23486054e-02 -1.10...
[ 2019-08-16 11:41:47,578 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,578 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 28  1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,578 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,578 ] [ DEBUG ] [ infer:130 ]  Partial infer for pre_rnn_conv/BiasAdd
[ 2019-08-16 11:41:47,578 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,578 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,579 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 28  1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,579 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [96], value = [-0.19918102 -0.17202342 -0.3115624   0.09023721 -0.37699068  0.01556218
 -0.45747772 -0.33676583...
[ 2019-08-16 11:41:47,579 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,579 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 28  1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,579 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,579 ] [ DEBUG ] [ infer:130 ]  Partial infer for pre_rnn_conv/LeakyRelu/mul
[ 2019-08-16 11:41:47,579 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,580 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,580 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0.2
[ 2019-08-16 11:41:47,580 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 28  1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,580 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,580 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 28  1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,580 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,580 ] [ DEBUG ] [ infer:130 ]  Partial infer for pre_rnn_conv/LeakyRelu
[ 2019-08-16 11:41:47,580 ] [ DEBUG ] [ infer:131 ]  Op: Maximum
[ 2019-08-16 11:41:47,580 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,580 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 28  1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,580 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 28  1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,580 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,581 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 28  1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,581 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,581 ] [ DEBUG ] [ infer:130 ]  Partial infer for reshape_to_rnn
[ 2019-08-16 11:41:47,581 ] [ DEBUG ] [ infer:131 ]  Op: Reshape
[ 2019-08-16 11:41:47,581 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,581 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 28  1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,581 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [3], value = [-1 28 96]
[ 2019-08-16 11:41:47,581 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,581 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 28 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,581 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,581 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/ReverseV2
[ 2019-08-16 11:41:47,582 ] [ DEBUG ] [ infer:131 ]  Op: Reverse
[ 2019-08-16 11:41:47,582 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,582 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 28 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,582 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,582 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 28 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,582 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,582 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/transpose
[ 2019-08-16 11:41:47,582 ] [ DEBUG ] [ infer:131 ]  Op: Transpose
[ 2019-08-16 11:41:47,583 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,583 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 28 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,583 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [3], value = [1 0 2]
[ 2019-08-16 11:41:47,583 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,583 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [28  1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,583 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,583 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArrayUnstack/Shape
[ 2019-08-16 11:41:47,583 ] [ DEBUG ] [ infer:131 ]  Op: ShapeOf
[ 2019-08-16 11:41:47,583 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,583 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [28  1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,583 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,583 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [3], value = [28  1 96]
[ 2019-08-16 11:41:47,584 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,584 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArrayUnstack/strided_slice
[ 2019-08-16 11:41:47,584 ] [ DEBUG ] [ infer:131 ]  Op: StridedSlice
[ 2019-08-16 11:41:47,584 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,585 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [3], value = [28  1 96]
[ 2019-08-16 11:41:47,585 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [0]
[ 2019-08-16 11:41:47,585 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,585 ] [ DEBUG ] [ infer:32 ]  input[3]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,585 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,585 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,585 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,585 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArrayUnstack/range
[ 2019-08-16 11:41:47,585 ] [ DEBUG ] [ infer:131 ]  Op: Range
[ 2019-08-16 11:41:47,585 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,585 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,585 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 28
[ 2019-08-16 11:41:47,585 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [], value = 1
[ 2019-08-16 11:41:47,586 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,586 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [28], value = [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
 24 25 26 27]
[ 2019-08-16 11:41:47,586 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,586 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/Shape_1
[ 2019-08-16 11:41:47,586 ] [ DEBUG ] [ infer:131 ]  Op: ShapeOf
[ 2019-08-16 11:41:47,586 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,586 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [28  1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,586 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,586 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [3], value = [28  1 96]
[ 2019-08-16 11:41:47,586 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,586 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/strided_slice_1
[ 2019-08-16 11:41:47,586 ] [ DEBUG ] [ infer:131 ]  Op: StridedSlice
[ 2019-08-16 11:41:47,587 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,587 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [3], value = [28  1 96]
[ 2019-08-16 11:41:47,587 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [0]
[ 2019-08-16 11:41:47,587 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,588 ] [ DEBUG ] [ infer:32 ]  input[3]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,588 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,588 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,588 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,588 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Less/Enter
[ 2019-08-16 11:41:47,588 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,588 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,588 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,588 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,588 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,588 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,588 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Less
[ 2019-08-16 11:41:47,588 ] [ DEBUG ] [ infer:131 ]  Op: Less
[ 2019-08-16 11:41:47,588 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,588 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,588 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 28
[ 2019-08-16 11:41:47,588 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,589 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,589 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,589 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/Maximum
[ 2019-08-16 11:41:47,589 ] [ DEBUG ] [ infer:131 ]  Op: Maximum
[ 2019-08-16 11:41:47,589 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,589 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,589 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 28
[ 2019-08-16 11:41:47,589 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,589 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,589 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,589 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/Minimum
[ 2019-08-16 11:41:47,589 ] [ DEBUG ] [ infer:131 ]  Op: Minimum
[ 2019-08-16 11:41:47,589 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,589 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,590 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 28
[ 2019-08-16 11:41:47,590 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,590 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,590 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,590 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Less_1/Enter
[ 2019-08-16 11:41:47,590 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,590 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,590 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,590 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,590 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,590 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,590 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Less_1
[ 2019-08-16 11:41:47,590 ] [ DEBUG ] [ infer:131 ]  Op: Less
[ 2019-08-16 11:41:47,590 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,590 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,590 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 28
[ 2019-08-16 11:41:47,590 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,591 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,591 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,591 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/LogicalAnd
[ 2019-08-16 11:41:47,591 ] [ DEBUG ] [ infer:131 ]  Op: LogicalAnd
[ 2019-08-16 11:41:47,591 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,591 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,591 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,591 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,591 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,591 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,591 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/LoopCond
[ 2019-08-16 11:41:47,591 ] [ DEBUG ] [ infer:131 ]  Op: LoopCond
[ 2019-08-16 11:41:47,591 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,591 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,591 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,591 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,591 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,592 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Switch_1
[ 2019-08-16 11:41:47,592 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,592 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,592 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,592 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,592 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,592 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,592 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,592 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Identity_1
[ 2019-08-16 11:41:47,592 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,592 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,592 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,592 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,592 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,592 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,592 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Switch
[ 2019-08-16 11:41:47,592 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,593 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,593 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,593 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,593 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,593 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,593 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,593 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Identity
[ 2019-08-16 11:41:47,593 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,593 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,593 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,593 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,593 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,593 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,593 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/add_1/y
[ 2019-08-16 11:41:47,593 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,593 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,594 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,594 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,594 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,594 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/add_1
[ 2019-08-16 11:41:47,594 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,594 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,594 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,594 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 1
[ 2019-08-16 11:41:47,594 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,594 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,594 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,594 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/NextIteration_1
[ 2019-08-16 11:41:47,594 ] [ DEBUG ] [ infer:131 ]  Op: NextIteration
[ 2019-08-16 11:41:47,594 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,594 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,594 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,594 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,595 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,595 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/sub/x
[ 2019-08-16 11:41:47,595 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,595 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,595 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,595 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1.0
[ 2019-08-16 11:41:47,595 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,595 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/concat_1/axis
[ 2019-08-16 11:41:47,595 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,595 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,595 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,595 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,595 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,595 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/split/split_dim
[ 2019-08-16 11:41:47,595 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,595 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,595 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,596 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,596 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,596 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/concat/axis
[ 2019-08-16 11:41:47,596 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,596 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,596 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,596 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,596 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,596 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/add/y
[ 2019-08-16 11:41:47,596 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,596 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,596 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,596 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,596 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,596 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/add
[ 2019-08-16 11:41:47,596 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,596 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,597 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,597 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 1
[ 2019-08-16 11:41:47,597 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,597 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,597 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,597 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/NextIteration
[ 2019-08-16 11:41:47,597 ] [ DEBUG ] [ infer:131 ]  Op: NextIteration
[ 2019-08-16 11:41:47,597 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,597 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,597 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,597 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,597 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,597 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArray_1
[ 2019-08-16 11:41:47,597 ] [ DEBUG ] [ infer:131 ]  Op: TensorArrayV3
[ 2019-08-16 11:41:47,597 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,597 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,597 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,598 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArray_1/Output_0/Data_
[ 2019-08-16 11:41:47,598 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,598 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,598 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/TensorArrayReadV3/Enter
[ 2019-08-16 11:41:47,598 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,598 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,598 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArray_1/Output_0/Data_
[ 2019-08-16 11:41:47,598 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,598 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArray_1/Output_0/Data_
[ 2019-08-16 11:41:47,598 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,598 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArrayUnstack/TensorArrayScatter/TensorArrayScatterV3
[ 2019-08-16 11:41:47,598 ] [ DEBUG ] [ infer:131 ]  Op: TensorArrayScatterV3
[ 2019-08-16 11:41:47,598 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,598 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArray_1/Output_0/Data_
[ 2019-08-16 11:41:47,599 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [28], value = [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
 24 25 26 27]
[ 2019-08-16 11:41:47,599 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [28  1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,599 ] [ DEBUG ] [ infer:32 ]  input[3]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,599 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,599 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,599 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,599 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/TensorArrayReadV3/Enter_1
[ 2019-08-16 11:41:47,599 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,599 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,599 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,599 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,599 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,599 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,599 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/TensorArrayReadV3
[ 2019-08-16 11:41:47,599 ] [ DEBUG ] [ infer:131 ]  Op: TensorArrayReadV3
[ 2019-08-16 11:41:47,600 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,600 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArray_1/Output_0/Data_
[ 2019-08-16 11:41:47,600 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,600 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,600 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,600 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,600 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,600 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArray
[ 2019-08-16 11:41:47,600 ] [ DEBUG ] [ infer:131 ]  Op: TensorArrayV3
[ 2019-08-16 11:41:47,600 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,600 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,600 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,600 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArray/Output_0/Data_
[ 2019-08-16 11:41:47,600 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,600 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,601 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/TensorArrayWrite/TensorArrayWriteV3/Enter
[ 2019-08-16 11:41:47,601 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,601 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,601 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArray/Output_0/Data_
[ 2019-08-16 11:41:47,601 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,601 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArray/Output_0/Data_
[ 2019-08-16 11:41:47,601 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,601 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Enter_2
[ 2019-08-16 11:41:47,601 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,601 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,601 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,601 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,601 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,601 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,601 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Merge_2
[ 2019-08-16 11:41:47,601 ] [ DEBUG ] [ infer:131 ]  Op: Merge
[ 2019-08-16 11:41:47,602 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,602 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,602 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = <UNKNOWN>, value = <UNKNOWN>
[ 2019-08-16 11:41:47,602 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,602 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,602 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,602 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Switch_2
[ 2019-08-16 11:41:47,602 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,602 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,602 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,602 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,602 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,602 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,602 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,602 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,603 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Identity_2
[ 2019-08-16 11:41:47,603 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,603 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,603 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,603 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,603 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,603 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,603 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/Shape
[ 2019-08-16 11:41:47,603 ] [ DEBUG ] [ infer:131 ]  Op: ShapeOf
[ 2019-08-16 11:41:47,603 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,603 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [28  1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,603 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,603 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [3], value = [28  1 96]
[ 2019-08-16 11:41:47,603 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,604 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/strided_slice
[ 2019-08-16 11:41:47,604 ] [ DEBUG ] [ infer:131 ]  Op: StridedSlice
[ 2019-08-16 11:41:47,604 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,604 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [3], value = [28  1 96]
[ 2019-08-16 11:41:47,604 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,605 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [1], value = [2]
[ 2019-08-16 11:41:47,605 ] [ DEBUG ] [ infer:32 ]  input[3]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,605 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,605 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,605 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,605 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/GRUCellZeroState/ExpandDims
[ 2019-08-16 11:41:47,605 ] [ DEBUG ] [ infer:131 ]  Op: Unsqueeze
[ 2019-08-16 11:41:47,606 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,606 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,606 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [0]
[ 2019-08-16 11:41:47,606 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,606 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,606 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,606 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/GRUCellZeroState/concat
[ 2019-08-16 11:41:47,606 ] [ DEBUG ] [ infer:131 ]  Op: ConcatV2
[ 2019-08-16 11:41:47,606 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,606 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,607 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [96]
[ 2019-08-16 11:41:47,607 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,607 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [2], value = [ 1 96]
[ 2019-08-16 11:41:47,607 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,607 ] [ DEBUG ] [ infer:130 ]  Partial infer for 770
[ 2019-08-16 11:41:47,607 ] [ DEBUG ] [ infer:131 ]  Op: Broadcast
[ 2019-08-16 11:41:47,607 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,607 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0.0
[ 2019-08-16 11:41:47,607 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [2], value = [ 1 96]
[ 2019-08-16 11:41:47,607 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,608 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = [[0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
  0. 0. 0. 0. 0. 0. 0. ...
[ 2019-08-16 11:41:47,608 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,608 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Enter_3
[ 2019-08-16 11:41:47,608 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,608 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,609 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = [[0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
  0. 0. 0. 0. 0. 0. 0. ...
[ 2019-08-16 11:41:47,609 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,609 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = [[0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
  0. 0. 0. 0. 0. 0. 0. ...
[ 2019-08-16 11:41:47,609 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,609 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Merge_3
[ 2019-08-16 11:41:47,610 ] [ DEBUG ] [ infer:131 ]  Op: Merge
[ 2019-08-16 11:41:47,610 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,610 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = [[0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
  0. 0. 0. 0. 0. 0. 0. ...
[ 2019-08-16 11:41:47,610 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = <UNKNOWN>, value = <UNKNOWN>
[ 2019-08-16 11:41:47,610 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,610 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,610 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,611 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Switch_3
[ 2019-08-16 11:41:47,611 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,611 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,611 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,611 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,611 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,611 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,611 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,611 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Identity_3
[ 2019-08-16 11:41:47,611 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,611 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,611 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,611 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,611 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,612 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,612 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/concat
[ 2019-08-16 11:41:47,612 ] [ DEBUG ] [ infer:131 ]  Op: ConcatV2
[ 2019-08-16 11:41:47,612 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,612 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,612 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,612 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,612 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,612 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,612 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/MatMul
[ 2019-08-16 11:41:47,612 ] [ DEBUG ] [ infer:131 ]  Op: MatMul
[ 2019-08-16 11:41:47,613 ] [ DEBUG ] [ matmul:30 ]  matmul shapes: [array([  1, 192]), array([192, 192])]
[ 2019-08-16 11:41:47,613 ] [ DEBUG ] [ matmul:63 ]  shape_tuple: (array([1]), array([192]))
[ 2019-08-16 11:41:47,613 ] [ DEBUG ] [ matmul:66 ]  matmul shape: [  1 192]
[ 2019-08-16 11:41:47,613 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,613 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,613 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [192 192], value = [[-0.18263446  0.31629124  0.73795736 ... -0.0191342  -0.1825211
  -0.31635457]
 [-0.23331062 -0....
[ 2019-08-16 11:41:47,613 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,613 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,614 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,614 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/BiasAdd
[ 2019-08-16 11:41:47,614 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,614 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,614 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,615 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [192], value = [0.6209333  0.5225718  0.3840976  1.0650713  0.8585766  0.8721105
 0.8514211  0.62742203 0.872603...
[ 2019-08-16 11:41:47,615 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,615 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,615 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,615 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/Sigmoid
[ 2019-08-16 11:41:47,615 ] [ DEBUG ] [ infer:131 ]  Op: Sigmoid
[ 2019-08-16 11:41:47,615 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,615 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,616 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,616 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,616 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,616 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/split
[ 2019-08-16 11:41:47,616 ] [ DEBUG ] [ infer:131 ]  Op: Split
[ 2019-08-16 11:41:47,616 ] [ DEBUG ] [ split:140 ]  input shape for split: [  1 192], should be split along 1 dim
[ 2019-08-16 11:41:47,616 ] [ DEBUG ] [ split:142 ]  split_dim_size type = <class 'numpy.int64'>
[ 2019-08-16 11:41:47,616 ] [ DEBUG ] [ split:150 ]  split_dim_size = 192, node.num_split = 2, div = 96.0, typeof div = <class 'numpy.float64'>
[ WARNING ]  Removed "axis" attribute from the scope of the model relayout pass because len(input.shape) == 2 != 4 for node rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/split
[ 2019-08-16 11:41:47,616 ] [ DEBUG ] [ split:118 ]  output shapes after split: [array([ 1, 96]), array([ 1, 96])]
[ 2019-08-16 11:41:47,616 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,616 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,617 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,617 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,617 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,617 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,617 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/sub/negate_
[ 2019-08-16 11:41:47,617 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,617 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,617 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,617 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = -1
[ 2019-08-16 11:41:47,617 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,617 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,617 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,617 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/sub/add_
[ 2019-08-16 11:41:47,617 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,618 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,618 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,618 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 1.0
[ 2019-08-16 11:41:47,618 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,618 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,618 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,618 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/mul_1
[ 2019-08-16 11:41:47,618 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,618 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,618 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,618 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,618 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,618 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,619 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,619 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/mul
[ 2019-08-16 11:41:47,619 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,619 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,619 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,619 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,619 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,619 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,619 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,619 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/concat_1
[ 2019-08-16 11:41:47,619 ] [ DEBUG ] [ infer:131 ]  Op: ConcatV2
[ 2019-08-16 11:41:47,619 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,620 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,620 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,620 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,620 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,620 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,620 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/MatMul_1
[ 2019-08-16 11:41:47,620 ] [ DEBUG ] [ infer:131 ]  Op: MatMul
[ 2019-08-16 11:41:47,620 ] [ DEBUG ] [ matmul:30 ]  matmul shapes: [array([  1, 192]), array([192,  96])]
[ 2019-08-16 11:41:47,620 ] [ DEBUG ] [ matmul:63 ]  shape_tuple: (array([1]), array([96]))
[ 2019-08-16 11:41:47,620 ] [ DEBUG ] [ matmul:66 ]  matmul shape: [ 1 96]
[ 2019-08-16 11:41:47,620 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,621 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,621 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [192  96], value = [[ 0.44494498  0.0179856   0.1610958  ... -0.42989033 -0.02577065
   0.42928106]
 [-0.0954172   0...
[ 2019-08-16 11:41:47,621 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,621 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,621 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,621 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/BiasAdd_1
[ 2019-08-16 11:41:47,621 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,621 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,621 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,622 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [96], value = [ 0.14719655  0.12150294 -0.14337113 -0.34720546  0.16993123  0.01489505
  0.00510493 -0.30431303...
[ 2019-08-16 11:41:47,622 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,622 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,622 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,622 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/Tanh
[ 2019-08-16 11:41:47,622 ] [ DEBUG ] [ infer:131 ]  Op: Tanh
[ 2019-08-16 11:41:47,623 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,623 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,623 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,623 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,623 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,623 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/mul_2
[ 2019-08-16 11:41:47,623 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,623 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,623 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,623 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,623 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,623 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,623 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,623 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/add
[ 2019-08-16 11:41:47,624 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,624 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,624 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,624 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,624 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,624 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,624 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,624 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/NextIteration_3
[ 2019-08-16 11:41:47,624 ] [ DEBUG ] [ infer:131 ]  Op: NextIteration
[ 2019-08-16 11:41:47,624 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,624 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,624 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,624 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,625 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,625 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/TensorArrayWrite/TensorArrayWriteV3
[ 2019-08-16 11:41:47,625 ] [ DEBUG ] [ infer:131 ]  Op: TensorArrayWriteV3
[ 2019-08-16 11:41:47,625 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,625 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArray/Output_0/Data_
[ 2019-08-16 11:41:47,625 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,625 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,625 ] [ DEBUG ] [ infer:32 ]  input[3]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,625 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,625 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,625 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,625 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/NextIteration_2
[ 2019-08-16 11:41:47,625 ] [ DEBUG ] [ infer:131 ]  Op: NextIteration
[ 2019-08-16 11:41:47,625 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,625 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,625 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,626 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,626 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,626 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Exit_2
[ 2019-08-16 11:41:47,626 ] [ DEBUG ] [ infer:131 ]  Op: Exit
[ 2019-08-16 11:41:47,626 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,626 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,626 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,626 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,626 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,626 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArrayStack/TensorArraySizeV3
[ 2019-08-16 11:41:47,626 ] [ DEBUG ] [ infer:131 ]  Op: TensorArraySizeV3
[ 2019-08-16 11:41:47,626 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,626 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArray/Output_0/Data_
[ 2019-08-16 11:41:47,626 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,626 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,626 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,627 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,627 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArrayStack/range
[ 2019-08-16 11:41:47,627 ] [ DEBUG ] [ infer:131 ]  Op: Range
[ 2019-08-16 11:41:47,627 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,627 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,627 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 28
[ 2019-08-16 11:41:47,627 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [], value = 1
[ 2019-08-16 11:41:47,627 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,627 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [28], value = [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
 24 25 26 27]
[ 2019-08-16 11:41:47,627 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,627 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArrayStack/TensorArrayGatherV3
[ 2019-08-16 11:41:47,627 ] [ DEBUG ] [ infer:131 ]  Op: TensorArrayGatherV3
[ 2019-08-16 11:41:47,627 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,628 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArray/Output_0/Data_
[ 2019-08-16 11:41:47,628 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [28], value = [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
 24 25 26 27]
[ 2019-08-16 11:41:47,628 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,628 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,628 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [28  1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,628 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,628 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/transpose_1
[ 2019-08-16 11:41:47,628 ] [ DEBUG ] [ infer:131 ]  Op: Transpose
[ 2019-08-16 11:41:47,629 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,629 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [28  1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,629 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [3], value = [1 0 2]
[ 2019-08-16 11:41:47,629 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,629 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 28 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,629 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,629 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/ReverseV2
[ 2019-08-16 11:41:47,629 ] [ DEBUG ] [ infer:131 ]  Op: Reverse
[ 2019-08-16 11:41:47,629 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,629 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 28 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,629 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,630 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 28 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,630 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,630 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/transpose
[ 2019-08-16 11:41:47,630 ] [ DEBUG ] [ infer:131 ]  Op: Transpose
[ 2019-08-16 11:41:47,630 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,630 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 28 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,630 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [3], value = [1 0 2]
[ 2019-08-16 11:41:47,630 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,630 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [28  1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,631 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,631 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArrayUnstack/Shape
[ 2019-08-16 11:41:47,631 ] [ DEBUG ] [ infer:131 ]  Op: ShapeOf
[ 2019-08-16 11:41:47,631 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,631 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [28  1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,631 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,631 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [3], value = [28  1 96]
[ 2019-08-16 11:41:47,631 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,631 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArrayUnstack/strided_slice
[ 2019-08-16 11:41:47,631 ] [ DEBUG ] [ infer:131 ]  Op: StridedSlice
[ 2019-08-16 11:41:47,632 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,632 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [3], value = [28  1 96]
[ 2019-08-16 11:41:47,632 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [0]
[ 2019-08-16 11:41:47,632 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,632 ] [ DEBUG ] [ infer:32 ]  input[3]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,632 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,632 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,633 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,633 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArrayUnstack/range
[ 2019-08-16 11:41:47,633 ] [ DEBUG ] [ infer:131 ]  Op: Range
[ 2019-08-16 11:41:47,633 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,633 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,633 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 28
[ 2019-08-16 11:41:47,633 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [], value = 1
[ 2019-08-16 11:41:47,633 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,633 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [28], value = [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
 24 25 26 27]
[ 2019-08-16 11:41:47,633 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,633 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/Shape_1
[ 2019-08-16 11:41:47,633 ] [ DEBUG ] [ infer:131 ]  Op: ShapeOf
[ 2019-08-16 11:41:47,634 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,634 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [28  1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,634 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,634 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [3], value = [28  1 96]
[ 2019-08-16 11:41:47,634 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,634 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/strided_slice_1
[ 2019-08-16 11:41:47,634 ] [ DEBUG ] [ infer:131 ]  Op: StridedSlice
[ 2019-08-16 11:41:47,635 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,635 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [3], value = [28  1 96]
[ 2019-08-16 11:41:47,635 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [0]
[ 2019-08-16 11:41:47,635 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,635 ] [ DEBUG ] [ infer:32 ]  input[3]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,635 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,635 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,635 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,635 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Less/Enter
[ 2019-08-16 11:41:47,635 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,635 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,635 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,636 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,636 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,636 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,636 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Less
[ 2019-08-16 11:41:47,636 ] [ DEBUG ] [ infer:131 ]  Op: Less
[ 2019-08-16 11:41:47,636 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,636 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,636 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 28
[ 2019-08-16 11:41:47,636 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,636 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,636 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,636 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/Maximum
[ 2019-08-16 11:41:47,636 ] [ DEBUG ] [ infer:131 ]  Op: Maximum
[ 2019-08-16 11:41:47,636 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,636 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,636 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 28
[ 2019-08-16 11:41:47,637 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,637 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,637 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,637 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/Minimum
[ 2019-08-16 11:41:47,637 ] [ DEBUG ] [ infer:131 ]  Op: Minimum
[ 2019-08-16 11:41:47,637 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,637 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,637 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 28
[ 2019-08-16 11:41:47,637 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,637 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,637 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,637 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Less_1/Enter
[ 2019-08-16 11:41:47,637 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,637 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,637 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,637 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,637 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,638 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,638 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Less_1
[ 2019-08-16 11:41:47,638 ] [ DEBUG ] [ infer:131 ]  Op: Less
[ 2019-08-16 11:41:47,638 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,638 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,638 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 28
[ 2019-08-16 11:41:47,638 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,638 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,638 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,638 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/LogicalAnd
[ 2019-08-16 11:41:47,638 ] [ DEBUG ] [ infer:131 ]  Op: LogicalAnd
[ 2019-08-16 11:41:47,638 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,638 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,638 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,638 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,638 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,639 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,639 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/LoopCond
[ 2019-08-16 11:41:47,639 ] [ DEBUG ] [ infer:131 ]  Op: LoopCond
[ 2019-08-16 11:41:47,639 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,639 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,639 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,639 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,639 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,639 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Switch_1
[ 2019-08-16 11:41:47,639 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,639 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,639 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,639 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,639 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,639 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,640 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,640 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Identity_1
[ 2019-08-16 11:41:47,640 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,640 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,640 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,640 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,640 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,640 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,640 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Switch
[ 2019-08-16 11:41:47,640 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,640 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,640 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,640 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,640 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,640 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,640 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,640 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Identity
[ 2019-08-16 11:41:47,641 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,641 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,641 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,641 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,641 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,641 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,641 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/add_1/y
[ 2019-08-16 11:41:47,641 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,641 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,641 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,641 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,641 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,641 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/add_1
[ 2019-08-16 11:41:47,641 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,641 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,641 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,641 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 1
[ 2019-08-16 11:41:47,642 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,642 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,642 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,642 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/NextIteration_1
[ 2019-08-16 11:41:47,642 ] [ DEBUG ] [ infer:131 ]  Op: NextIteration
[ 2019-08-16 11:41:47,642 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,642 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,642 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,642 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,642 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,642 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/sub/x
[ 2019-08-16 11:41:47,642 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,642 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,642 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,642 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1.0
[ 2019-08-16 11:41:47,642 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,642 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/concat_1/axis
[ 2019-08-16 11:41:47,642 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,643 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,643 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,643 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,643 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,643 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/split/split_dim
[ 2019-08-16 11:41:47,643 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,643 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,643 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,643 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,643 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,643 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/concat/axis
[ 2019-08-16 11:41:47,643 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,643 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,643 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,643 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,643 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,643 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/add/y
[ 2019-08-16 11:41:47,644 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,644 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,644 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,644 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,644 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,644 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/add
[ 2019-08-16 11:41:47,644 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,644 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,644 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,644 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 1
[ 2019-08-16 11:41:47,644 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,644 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,644 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,644 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/NextIteration
[ 2019-08-16 11:41:47,644 ] [ DEBUG ] [ infer:131 ]  Op: NextIteration
[ 2019-08-16 11:41:47,644 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,645 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,645 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,645 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,645 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,645 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArray_1
[ 2019-08-16 11:41:47,645 ] [ DEBUG ] [ infer:131 ]  Op: TensorArrayV3
[ 2019-08-16 11:41:47,645 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,645 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,645 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,645 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArray_1/Output_0/Data_
[ 2019-08-16 11:41:47,645 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,645 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,645 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/TensorArrayReadV3/Enter
[ 2019-08-16 11:41:47,645 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,645 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,645 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArray_1/Output_0/Data_
[ 2019-08-16 11:41:47,646 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,646 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArray_1/Output_0/Data_
[ 2019-08-16 11:41:47,646 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,646 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArrayUnstack/TensorArrayScatter/TensorArrayScatterV3
[ 2019-08-16 11:41:47,646 ] [ DEBUG ] [ infer:131 ]  Op: TensorArrayScatterV3
[ 2019-08-16 11:41:47,646 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,646 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArray_1/Output_0/Data_
[ 2019-08-16 11:41:47,646 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [28], value = [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
 24 25 26 27]
[ 2019-08-16 11:41:47,646 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [28  1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,646 ] [ DEBUG ] [ infer:32 ]  input[3]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,646 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,646 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,646 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,646 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/TensorArrayReadV3/Enter_1
[ 2019-08-16 11:41:47,647 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,647 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,647 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,647 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,647 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,647 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,647 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/TensorArrayReadV3
[ 2019-08-16 11:41:47,647 ] [ DEBUG ] [ infer:131 ]  Op: TensorArrayReadV3
[ 2019-08-16 11:41:47,647 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,647 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArray_1/Output_0/Data_
[ 2019-08-16 11:41:47,647 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,647 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,647 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,647 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,647 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,647 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArray
[ 2019-08-16 11:41:47,648 ] [ DEBUG ] [ infer:131 ]  Op: TensorArrayV3
[ 2019-08-16 11:41:47,648 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,648 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,648 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,648 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArray/Output_0/Data_
[ 2019-08-16 11:41:47,648 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,648 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,648 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/TensorArrayWrite/TensorArrayWriteV3/Enter
[ 2019-08-16 11:41:47,648 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,648 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,648 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArray/Output_0/Data_
[ 2019-08-16 11:41:47,648 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,648 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArray/Output_0/Data_
[ 2019-08-16 11:41:47,648 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,648 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Enter_2
[ 2019-08-16 11:41:47,648 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,649 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,649 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,649 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,649 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,649 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,649 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Merge_2
[ 2019-08-16 11:41:47,649 ] [ DEBUG ] [ infer:131 ]  Op: Merge
[ 2019-08-16 11:41:47,649 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,649 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,649 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = <UNKNOWN>, value = <UNKNOWN>
[ 2019-08-16 11:41:47,649 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,649 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,649 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,649 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Switch_2
[ 2019-08-16 11:41:47,649 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,649 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,650 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,650 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,650 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,650 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,650 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,650 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,650 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Identity_2
[ 2019-08-16 11:41:47,650 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,650 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,650 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,650 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,650 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,650 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,650 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/Shape
[ 2019-08-16 11:41:47,650 ] [ DEBUG ] [ infer:131 ]  Op: ShapeOf
[ 2019-08-16 11:41:47,651 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,651 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [28  1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,651 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,651 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [3], value = [28  1 96]
[ 2019-08-16 11:41:47,651 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,651 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/strided_slice
[ 2019-08-16 11:41:47,651 ] [ DEBUG ] [ infer:131 ]  Op: StridedSlice
[ 2019-08-16 11:41:47,652 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,652 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [3], value = [28  1 96]
[ 2019-08-16 11:41:47,652 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,652 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [1], value = [2]
[ 2019-08-16 11:41:47,652 ] [ DEBUG ] [ infer:32 ]  input[3]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,652 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,652 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,652 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,652 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/GRUCellZeroState/ExpandDims
[ 2019-08-16 11:41:47,652 ] [ DEBUG ] [ infer:131 ]  Op: Unsqueeze
[ 2019-08-16 11:41:47,653 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,653 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,653 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [0]
[ 2019-08-16 11:41:47,653 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,653 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,653 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,653 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/GRUCellZeroState/concat
[ 2019-08-16 11:41:47,653 ] [ DEBUG ] [ infer:131 ]  Op: ConcatV2
[ 2019-08-16 11:41:47,654 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,654 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,654 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [96]
[ 2019-08-16 11:41:47,654 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,654 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [2], value = [ 1 96]
[ 2019-08-16 11:41:47,654 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,654 ] [ DEBUG ] [ infer:130 ]  Partial infer for 769
[ 2019-08-16 11:41:47,654 ] [ DEBUG ] [ infer:131 ]  Op: Broadcast
[ 2019-08-16 11:41:47,654 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,655 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0.0
[ 2019-08-16 11:41:47,655 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [2], value = [ 1 96]
[ 2019-08-16 11:41:47,655 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,655 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = [[0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
  0. 0. 0. 0. 0. 0. 0. ...
[ 2019-08-16 11:41:47,655 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,655 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Enter_3
[ 2019-08-16 11:41:47,656 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,656 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,656 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = [[0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
  0. 0. 0. 0. 0. 0. 0. ...
[ 2019-08-16 11:41:47,656 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,657 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = [[0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
  0. 0. 0. 0. 0. 0. 0. ...
[ 2019-08-16 11:41:47,657 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,657 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Merge_3
[ 2019-08-16 11:41:47,657 ] [ DEBUG ] [ infer:131 ]  Op: Merge
[ 2019-08-16 11:41:47,657 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,658 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = [[0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
  0. 0. 0. 0. 0. 0. 0. ...
[ 2019-08-16 11:41:47,658 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = <UNKNOWN>, value = <UNKNOWN>
[ 2019-08-16 11:41:47,658 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,658 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,658 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,658 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Switch_3
[ 2019-08-16 11:41:47,658 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,658 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,658 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,658 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,658 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,658 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,659 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,659 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Identity_3
[ 2019-08-16 11:41:47,659 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,659 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,659 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,659 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,659 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,659 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,659 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/concat
[ 2019-08-16 11:41:47,659 ] [ DEBUG ] [ infer:131 ]  Op: ConcatV2
[ 2019-08-16 11:41:47,659 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,659 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,660 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,660 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,660 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,660 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,660 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/MatMul
[ 2019-08-16 11:41:47,660 ] [ DEBUG ] [ infer:131 ]  Op: MatMul
[ 2019-08-16 11:41:47,660 ] [ DEBUG ] [ matmul:30 ]  matmul shapes: [array([  1, 192]), array([192, 192])]
[ 2019-08-16 11:41:47,660 ] [ DEBUG ] [ matmul:63 ]  shape_tuple: (array([1]), array([192]))
[ 2019-08-16 11:41:47,660 ] [ DEBUG ] [ matmul:66 ]  matmul shape: [  1 192]
[ 2019-08-16 11:41:47,660 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,660 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,661 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [192 192], value = [[ 0.44174284 -0.10908651 -0.29394647 ... -0.13102415 -0.0380777
  -0.39950752]
 [ 0.2597118   0....
[ 2019-08-16 11:41:47,661 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,661 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,661 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,661 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/BiasAdd
[ 2019-08-16 11:41:47,661 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,661 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,661 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,662 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [192], value = [0.8748921  0.7196429  0.42538187 0.73728985 0.51834875 0.23964985
 0.68675274 0.54531384 0.40408...
[ 2019-08-16 11:41:47,662 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,663 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,663 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,663 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/Sigmoid
[ 2019-08-16 11:41:47,663 ] [ DEBUG ] [ infer:131 ]  Op: Sigmoid
[ 2019-08-16 11:41:47,663 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,663 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,663 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,663 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,663 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,663 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/split
[ 2019-08-16 11:41:47,663 ] [ DEBUG ] [ infer:131 ]  Op: Split
[ 2019-08-16 11:41:47,663 ] [ DEBUG ] [ split:140 ]  input shape for split: [  1 192], should be split along 1 dim
[ 2019-08-16 11:41:47,663 ] [ DEBUG ] [ split:142 ]  split_dim_size type = <class 'numpy.int64'>
[ 2019-08-16 11:41:47,663 ] [ DEBUG ] [ split:150 ]  split_dim_size = 192, node.num_split = 2, div = 96.0, typeof div = <class 'numpy.float64'>
[ WARNING ]  Removed "axis" attribute from the scope of the model relayout pass because len(input.shape) == 2 != 4 for node rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/split
[ 2019-08-16 11:41:47,664 ] [ DEBUG ] [ split:118 ]  output shapes after split: [array([ 1, 96]), array([ 1, 96])]
[ 2019-08-16 11:41:47,664 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,664 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,664 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,664 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,664 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,664 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,664 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/sub/negate_
[ 2019-08-16 11:41:47,664 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,664 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,665 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,665 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = -1
[ 2019-08-16 11:41:47,665 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,665 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,665 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,665 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/sub/add_
[ 2019-08-16 11:41:47,665 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,665 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,665 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,665 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 1.0
[ 2019-08-16 11:41:47,665 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,665 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,665 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,665 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/mul_1
[ 2019-08-16 11:41:47,665 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,666 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,666 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,666 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,666 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,666 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,666 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,666 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/mul
[ 2019-08-16 11:41:47,666 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,666 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,666 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,666 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,666 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,667 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,667 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,667 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/concat_1
[ 2019-08-16 11:41:47,667 ] [ DEBUG ] [ infer:131 ]  Op: ConcatV2
[ 2019-08-16 11:41:47,667 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,667 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,667 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,667 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,667 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,667 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,667 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/MatMul_1
[ 2019-08-16 11:41:47,667 ] [ DEBUG ] [ infer:131 ]  Op: MatMul
[ 2019-08-16 11:41:47,668 ] [ DEBUG ] [ matmul:30 ]  matmul shapes: [array([  1, 192]), array([192,  96])]
[ 2019-08-16 11:41:47,668 ] [ DEBUG ] [ matmul:63 ]  shape_tuple: (array([1]), array([96]))
[ 2019-08-16 11:41:47,668 ] [ DEBUG ] [ matmul:66 ]  matmul shape: [ 1 96]
[ 2019-08-16 11:41:47,668 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,668 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,668 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [192  96], value = [[ 0.07099124  0.03735324 -0.12896395 ...  0.0217509   0.10517395
  -0.06781325]
 [ 0.56132555 -0...
[ 2019-08-16 11:41:47,668 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,669 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,669 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,669 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/BiasAdd_1
[ 2019-08-16 11:41:47,669 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,669 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,669 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,670 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [96], value = [ 0.22845171 -0.02988362 -0.05550991  0.35331067 -0.11666836  0.02229408
  0.20225537  0.07914946...
[ 2019-08-16 11:41:47,670 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,670 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,670 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,670 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/Tanh
[ 2019-08-16 11:41:47,670 ] [ DEBUG ] [ infer:131 ]  Op: Tanh
[ 2019-08-16 11:41:47,670 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,670 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,670 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,670 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,670 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,670 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/mul_2
[ 2019-08-16 11:41:47,670 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,671 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,671 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,671 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,671 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,671 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,671 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,671 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/add
[ 2019-08-16 11:41:47,671 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,671 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,671 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,671 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,671 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,671 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,672 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,672 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/NextIteration_3
[ 2019-08-16 11:41:47,672 ] [ DEBUG ] [ infer:131 ]  Op: NextIteration
[ 2019-08-16 11:41:47,672 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,672 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,672 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,672 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,672 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,672 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/TensorArrayWrite/TensorArrayWriteV3
[ 2019-08-16 11:41:47,672 ] [ DEBUG ] [ infer:131 ]  Op: TensorArrayWriteV3
[ 2019-08-16 11:41:47,672 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,672 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArray/Output_0/Data_
[ 2019-08-16 11:41:47,672 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,672 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,672 ] [ DEBUG ] [ infer:32 ]  input[3]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,673 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,673 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,673 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,673 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/NextIteration_2
[ 2019-08-16 11:41:47,673 ] [ DEBUG ] [ infer:131 ]  Op: NextIteration
[ 2019-08-16 11:41:47,673 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,673 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,673 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,673 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,673 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,673 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Exit_2
[ 2019-08-16 11:41:47,673 ] [ DEBUG ] [ infer:131 ]  Op: Exit
[ 2019-08-16 11:41:47,673 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,673 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,673 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,673 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,673 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,674 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArrayStack/TensorArraySizeV3
[ 2019-08-16 11:41:47,674 ] [ DEBUG ] [ infer:131 ]  Op: TensorArraySizeV3
[ 2019-08-16 11:41:47,674 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,674 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArray/Output_0/Data_
[ 2019-08-16 11:41:47,674 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,674 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,674 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,674 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,674 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArrayStack/range
[ 2019-08-16 11:41:47,674 ] [ DEBUG ] [ infer:131 ]  Op: Range
[ 2019-08-16 11:41:47,674 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,674 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,674 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 28
[ 2019-08-16 11:41:47,674 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [], value = 1
[ 2019-08-16 11:41:47,674 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,675 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [28], value = [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
 24 25 26 27]
[ 2019-08-16 11:41:47,675 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,675 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArrayStack/TensorArrayGatherV3
[ 2019-08-16 11:41:47,675 ] [ DEBUG ] [ infer:131 ]  Op: TensorArrayGatherV3
[ 2019-08-16 11:41:47,675 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,675 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArray/Output_0/Data_
[ 2019-08-16 11:41:47,675 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [28], value = [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
 24 25 26 27]
[ 2019-08-16 11:41:47,675 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,675 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,675 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [28  1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,675 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,675 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/transpose_1
[ 2019-08-16 11:41:47,676 ] [ DEBUG ] [ infer:131 ]  Op: Transpose
[ 2019-08-16 11:41:47,676 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,676 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [28  1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,676 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [3], value = [1 0 2]
[ 2019-08-16 11:41:47,676 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,676 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 28 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,676 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,676 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_0/concat
[ 2019-08-16 11:41:47,677 ] [ DEBUG ] [ infer:131 ]  Op: ConcatV2
[ 2019-08-16 11:41:47,677 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,677 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 28 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,677 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 28 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,677 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,677 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1  28 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,677 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,677 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/ReverseV2
[ 2019-08-16 11:41:47,677 ] [ DEBUG ] [ infer:131 ]  Op: Reverse
[ 2019-08-16 11:41:47,677 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,678 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1  28 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,678 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,678 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1  28 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,678 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,678 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/transpose
[ 2019-08-16 11:41:47,678 ] [ DEBUG ] [ infer:131 ]  Op: Transpose
[ 2019-08-16 11:41:47,678 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,678 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1  28 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,678 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [3], value = [1 0 2]
[ 2019-08-16 11:41:47,678 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,679 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 28   1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,679 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,679 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArrayUnstack/Shape
[ 2019-08-16 11:41:47,679 ] [ DEBUG ] [ infer:131 ]  Op: ShapeOf
[ 2019-08-16 11:41:47,679 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,679 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 28   1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,679 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,679 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [3], value = [ 28   1 192]
[ 2019-08-16 11:41:47,679 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,679 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArrayUnstack/strided_slice
[ 2019-08-16 11:41:47,679 ] [ DEBUG ] [ infer:131 ]  Op: StridedSlice
[ 2019-08-16 11:41:47,680 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,680 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [3], value = [ 28   1 192]
[ 2019-08-16 11:41:47,680 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [0]
[ 2019-08-16 11:41:47,680 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,681 ] [ DEBUG ] [ infer:32 ]  input[3]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,681 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,681 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,681 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,681 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArrayUnstack/range
[ 2019-08-16 11:41:47,681 ] [ DEBUG ] [ infer:131 ]  Op: Range
[ 2019-08-16 11:41:47,681 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,681 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,681 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 28
[ 2019-08-16 11:41:47,681 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [], value = 1
[ 2019-08-16 11:41:47,681 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,681 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [28], value = [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
 24 25 26 27]
[ 2019-08-16 11:41:47,681 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,681 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/Shape_1
[ 2019-08-16 11:41:47,682 ] [ DEBUG ] [ infer:131 ]  Op: ShapeOf
[ 2019-08-16 11:41:47,682 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,682 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 28   1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,682 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,682 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [3], value = [ 28   1 192]
[ 2019-08-16 11:41:47,682 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,682 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/strided_slice_1
[ 2019-08-16 11:41:47,682 ] [ DEBUG ] [ infer:131 ]  Op: StridedSlice
[ 2019-08-16 11:41:47,683 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,683 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [3], value = [ 28   1 192]
[ 2019-08-16 11:41:47,683 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [0]
[ 2019-08-16 11:41:47,683 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,683 ] [ DEBUG ] [ infer:32 ]  input[3]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,683 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,683 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,683 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,683 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Less/Enter
[ 2019-08-16 11:41:47,683 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,684 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,684 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,684 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,684 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,684 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,684 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Less
[ 2019-08-16 11:41:47,684 ] [ DEBUG ] [ infer:131 ]  Op: Less
[ 2019-08-16 11:41:47,684 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,684 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,684 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 28
[ 2019-08-16 11:41:47,684 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,684 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,684 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,684 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/Maximum
[ 2019-08-16 11:41:47,684 ] [ DEBUG ] [ infer:131 ]  Op: Maximum
[ 2019-08-16 11:41:47,685 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,685 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,685 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 28
[ 2019-08-16 11:41:47,685 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,685 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,685 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,685 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/Minimum
[ 2019-08-16 11:41:47,685 ] [ DEBUG ] [ infer:131 ]  Op: Minimum
[ 2019-08-16 11:41:47,685 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,685 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,685 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 28
[ 2019-08-16 11:41:47,685 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,685 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,685 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,685 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Less_1/Enter
[ 2019-08-16 11:41:47,685 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,685 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,686 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,686 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,686 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,686 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,686 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Less_1
[ 2019-08-16 11:41:47,686 ] [ DEBUG ] [ infer:131 ]  Op: Less
[ 2019-08-16 11:41:47,686 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,686 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,686 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 28
[ 2019-08-16 11:41:47,686 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,686 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,686 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,686 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/LogicalAnd
[ 2019-08-16 11:41:47,686 ] [ DEBUG ] [ infer:131 ]  Op: LogicalAnd
[ 2019-08-16 11:41:47,686 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,687 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,687 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,687 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,687 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,687 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,687 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/LoopCond
[ 2019-08-16 11:41:47,687 ] [ DEBUG ] [ infer:131 ]  Op: LoopCond
[ 2019-08-16 11:41:47,687 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,687 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,687 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,687 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,687 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,687 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Switch_1
[ 2019-08-16 11:41:47,687 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,687 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,687 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,687 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,687 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,688 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,688 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,688 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Identity_1
[ 2019-08-16 11:41:47,688 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,688 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,688 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,688 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,688 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,688 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,688 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Switch
[ 2019-08-16 11:41:47,688 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,688 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,688 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,688 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,688 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,688 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,689 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,689 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Identity
[ 2019-08-16 11:41:47,689 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,689 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,689 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,689 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,689 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,689 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,689 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/add_1/y
[ 2019-08-16 11:41:47,689 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,689 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,689 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,689 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,689 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,689 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/add_1
[ 2019-08-16 11:41:47,689 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,689 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,690 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,690 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 1
[ 2019-08-16 11:41:47,690 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,690 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,690 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,690 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/NextIteration_1
[ 2019-08-16 11:41:47,690 ] [ DEBUG ] [ infer:131 ]  Op: NextIteration
[ 2019-08-16 11:41:47,690 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,690 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,690 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,690 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,690 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,690 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/sub/x
[ 2019-08-16 11:41:47,690 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,690 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,690 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,690 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1.0
[ 2019-08-16 11:41:47,690 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,691 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/concat_1/axis
[ 2019-08-16 11:41:47,691 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,691 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,691 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,691 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,691 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,691 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/split/split_dim
[ 2019-08-16 11:41:47,691 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,691 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,691 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,691 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,691 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,691 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/concat/axis
[ 2019-08-16 11:41:47,691 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,691 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,691 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,691 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,692 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,692 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/add/y
[ 2019-08-16 11:41:47,692 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,692 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,692 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,692 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,692 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,692 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/add
[ 2019-08-16 11:41:47,692 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,692 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,692 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,692 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 1
[ 2019-08-16 11:41:47,692 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,692 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,692 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,692 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/NextIteration
[ 2019-08-16 11:41:47,692 ] [ DEBUG ] [ infer:131 ]  Op: NextIteration
[ 2019-08-16 11:41:47,693 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,693 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,693 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,693 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,693 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,693 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArray_1
[ 2019-08-16 11:41:47,693 ] [ DEBUG ] [ infer:131 ]  Op: TensorArrayV3
[ 2019-08-16 11:41:47,693 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,693 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,693 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,693 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArray_1/Output_0/Data_
[ 2019-08-16 11:41:47,693 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,693 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,693 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/TensorArrayReadV3/Enter
[ 2019-08-16 11:41:47,693 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,693 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,694 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArray_1/Output_0/Data_
[ 2019-08-16 11:41:47,694 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,694 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArray_1/Output_0/Data_
[ 2019-08-16 11:41:47,694 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,694 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArrayUnstack/TensorArrayScatter/TensorArrayScatterV3
[ 2019-08-16 11:41:47,694 ] [ DEBUG ] [ infer:131 ]  Op: TensorArrayScatterV3
[ 2019-08-16 11:41:47,694 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,694 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArray_1/Output_0/Data_
[ 2019-08-16 11:41:47,694 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [28], value = [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
 24 25 26 27]
[ 2019-08-16 11:41:47,694 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [ 28   1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,694 ] [ DEBUG ] [ infer:32 ]  input[3]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,694 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,694 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,695 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,695 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/TensorArrayReadV3/Enter_1
[ 2019-08-16 11:41:47,695 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,695 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,695 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,695 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,695 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,695 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,695 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/TensorArrayReadV3
[ 2019-08-16 11:41:47,695 ] [ DEBUG ] [ infer:131 ]  Op: TensorArrayReadV3
[ 2019-08-16 11:41:47,695 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,695 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArray_1/Output_0/Data_
[ 2019-08-16 11:41:47,695 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,695 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,695 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,695 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,696 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,696 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArray
[ 2019-08-16 11:41:47,696 ] [ DEBUG ] [ infer:131 ]  Op: TensorArrayV3
[ 2019-08-16 11:41:47,696 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,696 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,696 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,696 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArray/Output_0/Data_
[ 2019-08-16 11:41:47,696 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,696 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,696 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/TensorArrayWrite/TensorArrayWriteV3/Enter
[ 2019-08-16 11:41:47,696 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,696 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,696 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArray/Output_0/Data_
[ 2019-08-16 11:41:47,696 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,696 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArray/Output_0/Data_
[ 2019-08-16 11:41:47,696 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,697 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Enter_2
[ 2019-08-16 11:41:47,697 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,697 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,697 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,697 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,697 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,697 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,697 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Merge_2
[ 2019-08-16 11:41:47,697 ] [ DEBUG ] [ infer:131 ]  Op: Merge
[ 2019-08-16 11:41:47,697 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,697 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,697 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = <UNKNOWN>, value = <UNKNOWN>
[ 2019-08-16 11:41:47,697 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,697 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,697 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,697 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Switch_2
[ 2019-08-16 11:41:47,697 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,698 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,698 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,698 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,698 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,698 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,698 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,698 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,698 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Identity_2
[ 2019-08-16 11:41:47,698 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,698 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,698 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,698 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,698 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,698 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,698 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/Shape
[ 2019-08-16 11:41:47,698 ] [ DEBUG ] [ infer:131 ]  Op: ShapeOf
[ 2019-08-16 11:41:47,699 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,699 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 28   1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,699 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,699 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [3], value = [ 28   1 192]
[ 2019-08-16 11:41:47,699 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,699 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/strided_slice
[ 2019-08-16 11:41:47,699 ] [ DEBUG ] [ infer:131 ]  Op: StridedSlice
[ 2019-08-16 11:41:47,700 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,700 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [3], value = [ 28   1 192]
[ 2019-08-16 11:41:47,700 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,700 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [1], value = [2]
[ 2019-08-16 11:41:47,700 ] [ DEBUG ] [ infer:32 ]  input[3]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,700 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,700 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,700 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,700 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/GRUCellZeroState/ExpandDims
[ 2019-08-16 11:41:47,700 ] [ DEBUG ] [ infer:131 ]  Op: Unsqueeze
[ 2019-08-16 11:41:47,701 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,701 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,701 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [0]
[ 2019-08-16 11:41:47,701 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,701 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,701 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,701 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/GRUCellZeroState/concat
[ 2019-08-16 11:41:47,701 ] [ DEBUG ] [ infer:131 ]  Op: ConcatV2
[ 2019-08-16 11:41:47,702 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,702 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,702 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [96]
[ 2019-08-16 11:41:47,702 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,702 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [2], value = [ 1 96]
[ 2019-08-16 11:41:47,702 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,702 ] [ DEBUG ] [ infer:130 ]  Partial infer for 772
[ 2019-08-16 11:41:47,702 ] [ DEBUG ] [ infer:131 ]  Op: Broadcast
[ 2019-08-16 11:41:47,703 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,703 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0.0
[ 2019-08-16 11:41:47,703 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [2], value = [ 1 96]
[ 2019-08-16 11:41:47,703 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,703 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = [[0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
  0. 0. 0. 0. 0. 0. 0. ...
[ 2019-08-16 11:41:47,703 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,704 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Enter_3
[ 2019-08-16 11:41:47,704 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,704 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,704 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = [[0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
  0. 0. 0. 0. 0. 0. 0. ...
[ 2019-08-16 11:41:47,704 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,705 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = [[0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
  0. 0. 0. 0. 0. 0. 0. ...
[ 2019-08-16 11:41:47,705 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,705 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Merge_3
[ 2019-08-16 11:41:47,705 ] [ DEBUG ] [ infer:131 ]  Op: Merge
[ 2019-08-16 11:41:47,705 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,706 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = [[0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
  0. 0. 0. 0. 0. 0. 0. ...
[ 2019-08-16 11:41:47,706 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = <UNKNOWN>, value = <UNKNOWN>
[ 2019-08-16 11:41:47,706 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,706 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,706 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,706 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Switch_3
[ 2019-08-16 11:41:47,706 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,706 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,706 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,706 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,706 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,706 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,707 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,707 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Identity_3
[ 2019-08-16 11:41:47,707 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,707 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,707 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,707 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,707 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,707 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,707 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/concat
[ 2019-08-16 11:41:47,707 ] [ DEBUG ] [ infer:131 ]  Op: ConcatV2
[ 2019-08-16 11:41:47,707 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,707 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,708 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,708 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,708 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 288], value = <UNKNOWN>
[ 2019-08-16 11:41:47,708 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,708 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/MatMul
[ 2019-08-16 11:41:47,708 ] [ DEBUG ] [ infer:131 ]  Op: MatMul
[ 2019-08-16 11:41:47,708 ] [ DEBUG ] [ matmul:30 ]  matmul shapes: [array([  1, 288]), array([288, 192])]
[ 2019-08-16 11:41:47,708 ] [ DEBUG ] [ matmul:63 ]  shape_tuple: (array([1]), array([192]))
[ 2019-08-16 11:41:47,708 ] [ DEBUG ] [ matmul:66 ]  matmul shape: [  1 192]
[ 2019-08-16 11:41:47,708 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,708 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 288], value = <UNKNOWN>
[ 2019-08-16 11:41:47,709 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [288 192], value = [[-1.0303181   0.29413357 -0.37989917 ...  0.24760938  0.15866868
   0.04757554]
 [-0.11078571 -0...
[ 2019-08-16 11:41:47,709 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,709 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,709 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,709 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/BiasAdd
[ 2019-08-16 11:41:47,709 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,709 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,709 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,710 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [192], value = [0.8636338  1.3107551  1.0014389  0.9294754  1.129363   0.976413
 0.6523678  0.8062803  0.8824739...
[ 2019-08-16 11:41:47,710 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,711 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,711 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,711 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/Sigmoid
[ 2019-08-16 11:41:47,711 ] [ DEBUG ] [ infer:131 ]  Op: Sigmoid
[ 2019-08-16 11:41:47,711 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,711 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,711 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,711 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,711 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,711 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/split
[ 2019-08-16 11:41:47,711 ] [ DEBUG ] [ infer:131 ]  Op: Split
[ 2019-08-16 11:41:47,711 ] [ DEBUG ] [ split:140 ]  input shape for split: [  1 192], should be split along 1 dim
[ 2019-08-16 11:41:47,711 ] [ DEBUG ] [ split:142 ]  split_dim_size type = <class 'numpy.int64'>
[ 2019-08-16 11:41:47,711 ] [ DEBUG ] [ split:150 ]  split_dim_size = 192, node.num_split = 2, div = 96.0, typeof div = <class 'numpy.float64'>
[ WARNING ]  Removed "axis" attribute from the scope of the model relayout pass because len(input.shape) == 2 != 4 for node rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/split
[ 2019-08-16 11:41:47,712 ] [ DEBUG ] [ split:118 ]  output shapes after split: [array([ 1, 96]), array([ 1, 96])]
[ 2019-08-16 11:41:47,712 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,712 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,712 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,712 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,712 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,712 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,712 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/sub/negate_
[ 2019-08-16 11:41:47,712 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,712 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,713 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,713 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = -1
[ 2019-08-16 11:41:47,713 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,713 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,713 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,713 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/sub/add_
[ 2019-08-16 11:41:47,713 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,713 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,713 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,713 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 1.0
[ 2019-08-16 11:41:47,713 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,713 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,713 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,713 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/mul_1
[ 2019-08-16 11:41:47,713 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,714 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,714 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,714 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,714 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,714 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,714 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,714 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/mul
[ 2019-08-16 11:41:47,714 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,714 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,714 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,714 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,714 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,715 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,715 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,715 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/concat_1
[ 2019-08-16 11:41:47,715 ] [ DEBUG ] [ infer:131 ]  Op: ConcatV2
[ 2019-08-16 11:41:47,715 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,715 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,715 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,715 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,715 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 288], value = <UNKNOWN>
[ 2019-08-16 11:41:47,715 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,715 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/MatMul_1
[ 2019-08-16 11:41:47,715 ] [ DEBUG ] [ infer:131 ]  Op: MatMul
[ 2019-08-16 11:41:47,716 ] [ DEBUG ] [ matmul:30 ]  matmul shapes: [array([  1, 288]), array([288,  96])]
[ 2019-08-16 11:41:47,716 ] [ DEBUG ] [ matmul:63 ]  shape_tuple: (array([1]), array([96]))
[ 2019-08-16 11:41:47,716 ] [ DEBUG ] [ matmul:66 ]  matmul shape: [ 1 96]
[ 2019-08-16 11:41:47,716 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,716 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 288], value = <UNKNOWN>
[ 2019-08-16 11:41:47,716 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [288  96], value = [[ 0.12273635 -0.07469086  0.13450307 ... -0.38907412  0.09408648
   0.09192951]
 [-0.03125457 -0...
[ 2019-08-16 11:41:47,716 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,716 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,717 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,717 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/BiasAdd_1
[ 2019-08-16 11:41:47,717 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,717 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,717 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,718 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [96], value = [-0.20630556 -0.2388258   0.32675105 -0.01013813  0.21144664 -0.11858168
  0.05351019  0.02979827...
[ 2019-08-16 11:41:47,718 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,718 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,718 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,718 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/Tanh
[ 2019-08-16 11:41:47,718 ] [ DEBUG ] [ infer:131 ]  Op: Tanh
[ 2019-08-16 11:41:47,718 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,718 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,718 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,718 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,718 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,718 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/mul_2
[ 2019-08-16 11:41:47,718 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,719 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,719 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,719 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,719 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,719 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,719 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,719 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/add
[ 2019-08-16 11:41:47,719 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,719 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,719 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,719 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,719 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,719 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,719 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,720 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/NextIteration_3
[ 2019-08-16 11:41:47,720 ] [ DEBUG ] [ infer:131 ]  Op: NextIteration
[ 2019-08-16 11:41:47,720 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,720 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,720 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,720 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,720 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,720 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/TensorArrayWrite/TensorArrayWriteV3
[ 2019-08-16 11:41:47,720 ] [ DEBUG ] [ infer:131 ]  Op: TensorArrayWriteV3
[ 2019-08-16 11:41:47,720 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,720 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArray/Output_0/Data_
[ 2019-08-16 11:41:47,720 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,720 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,720 ] [ DEBUG ] [ infer:32 ]  input[3]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,721 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,721 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,721 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,721 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/NextIteration_2
[ 2019-08-16 11:41:47,721 ] [ DEBUG ] [ infer:131 ]  Op: NextIteration
[ 2019-08-16 11:41:47,721 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,721 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,721 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,721 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,721 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,721 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Exit_2
[ 2019-08-16 11:41:47,721 ] [ DEBUG ] [ infer:131 ]  Op: Exit
[ 2019-08-16 11:41:47,721 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,721 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,721 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,721 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,721 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,722 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArrayStack/TensorArraySizeV3
[ 2019-08-16 11:41:47,722 ] [ DEBUG ] [ infer:131 ]  Op: TensorArraySizeV3
[ 2019-08-16 11:41:47,722 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,722 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArray/Output_0/Data_
[ 2019-08-16 11:41:47,722 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,722 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,722 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,722 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,722 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArrayStack/range
[ 2019-08-16 11:41:47,722 ] [ DEBUG ] [ infer:131 ]  Op: Range
[ 2019-08-16 11:41:47,722 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,722 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,722 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 28
[ 2019-08-16 11:41:47,722 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [], value = 1
[ 2019-08-16 11:41:47,722 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,723 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [28], value = [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
 24 25 26 27]
[ 2019-08-16 11:41:47,723 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,723 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArrayStack/TensorArrayGatherV3
[ 2019-08-16 11:41:47,723 ] [ DEBUG ] [ infer:131 ]  Op: TensorArrayGatherV3
[ 2019-08-16 11:41:47,723 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,723 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArray/Output_0/Data_
[ 2019-08-16 11:41:47,723 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [28], value = [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
 24 25 26 27]
[ 2019-08-16 11:41:47,723 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,723 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,723 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [28  1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,723 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,723 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/transpose_1
[ 2019-08-16 11:41:47,724 ] [ DEBUG ] [ infer:131 ]  Op: Transpose
[ 2019-08-16 11:41:47,724 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,724 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [28  1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,724 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [3], value = [1 0 2]
[ 2019-08-16 11:41:47,724 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,724 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 28 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,724 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,724 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/ReverseV2
[ 2019-08-16 11:41:47,725 ] [ DEBUG ] [ infer:131 ]  Op: Reverse
[ 2019-08-16 11:41:47,725 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,725 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 28 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,725 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,725 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 28 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,725 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,725 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/transpose
[ 2019-08-16 11:41:47,725 ] [ DEBUG ] [ infer:131 ]  Op: Transpose
[ 2019-08-16 11:41:47,726 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,726 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1  28 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,726 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [3], value = [1 0 2]
[ 2019-08-16 11:41:47,726 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,726 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 28   1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,726 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,726 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArrayUnstack/Shape
[ 2019-08-16 11:41:47,726 ] [ DEBUG ] [ infer:131 ]  Op: ShapeOf
[ 2019-08-16 11:41:47,726 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,726 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 28   1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,726 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,726 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [3], value = [ 28   1 192]
[ 2019-08-16 11:41:47,727 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,727 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArrayUnstack/strided_slice
[ 2019-08-16 11:41:47,727 ] [ DEBUG ] [ infer:131 ]  Op: StridedSlice
[ 2019-08-16 11:41:47,727 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,727 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [3], value = [ 28   1 192]
[ 2019-08-16 11:41:47,728 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [0]
[ 2019-08-16 11:41:47,728 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,728 ] [ DEBUG ] [ infer:32 ]  input[3]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,728 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,728 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,728 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,728 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArrayUnstack/range
[ 2019-08-16 11:41:47,728 ] [ DEBUG ] [ infer:131 ]  Op: Range
[ 2019-08-16 11:41:47,728 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,728 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,728 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 28
[ 2019-08-16 11:41:47,728 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [], value = 1
[ 2019-08-16 11:41:47,728 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,729 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [28], value = [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
 24 25 26 27]
[ 2019-08-16 11:41:47,729 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,729 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/Shape_1
[ 2019-08-16 11:41:47,729 ] [ DEBUG ] [ infer:131 ]  Op: ShapeOf
[ 2019-08-16 11:41:47,729 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,729 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 28   1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,729 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,729 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [3], value = [ 28   1 192]
[ 2019-08-16 11:41:47,729 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,729 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/strided_slice_1
[ 2019-08-16 11:41:47,729 ] [ DEBUG ] [ infer:131 ]  Op: StridedSlice
[ 2019-08-16 11:41:47,730 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,730 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [3], value = [ 28   1 192]
[ 2019-08-16 11:41:47,730 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [0]
[ 2019-08-16 11:41:47,731 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,731 ] [ DEBUG ] [ infer:32 ]  input[3]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,731 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,731 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,731 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,731 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Less/Enter
[ 2019-08-16 11:41:47,731 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,731 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,731 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,731 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,731 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,731 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,731 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Less
[ 2019-08-16 11:41:47,731 ] [ DEBUG ] [ infer:131 ]  Op: Less
[ 2019-08-16 11:41:47,732 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,732 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,732 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 28
[ 2019-08-16 11:41:47,732 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,732 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,732 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,732 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/Maximum
[ 2019-08-16 11:41:47,732 ] [ DEBUG ] [ infer:131 ]  Op: Maximum
[ 2019-08-16 11:41:47,732 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,732 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,732 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 28
[ 2019-08-16 11:41:47,732 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,732 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,732 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,732 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/Minimum
[ 2019-08-16 11:41:47,732 ] [ DEBUG ] [ infer:131 ]  Op: Minimum
[ 2019-08-16 11:41:47,733 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,733 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,733 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 28
[ 2019-08-16 11:41:47,733 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,733 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,733 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,733 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Less_1/Enter
[ 2019-08-16 11:41:47,733 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,733 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,733 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,733 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,733 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,733 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,733 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Less_1
[ 2019-08-16 11:41:47,733 ] [ DEBUG ] [ infer:131 ]  Op: Less
[ 2019-08-16 11:41:47,733 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,733 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,734 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 28
[ 2019-08-16 11:41:47,734 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,734 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,734 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,734 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/LogicalAnd
[ 2019-08-16 11:41:47,734 ] [ DEBUG ] [ infer:131 ]  Op: LogicalAnd
[ 2019-08-16 11:41:47,734 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,734 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,734 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,734 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,734 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,734 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,734 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/LoopCond
[ 2019-08-16 11:41:47,734 ] [ DEBUG ] [ infer:131 ]  Op: LoopCond
[ 2019-08-16 11:41:47,734 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,735 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,735 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,735 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,735 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,735 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Switch_1
[ 2019-08-16 11:41:47,735 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,735 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,735 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,735 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,735 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,735 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,735 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,735 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Identity_1
[ 2019-08-16 11:41:47,735 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,735 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,735 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,735 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,736 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,736 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,736 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Switch
[ 2019-08-16 11:41:47,736 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,736 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,736 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,736 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,736 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,736 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,736 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,736 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Identity
[ 2019-08-16 11:41:47,736 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,736 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,736 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,736 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,736 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,737 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,737 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/add_1/y
[ 2019-08-16 11:41:47,737 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,737 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,737 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,737 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,737 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,737 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/add_1
[ 2019-08-16 11:41:47,737 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,737 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,737 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,737 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 1
[ 2019-08-16 11:41:47,737 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,737 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,737 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,737 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/NextIteration_1
[ 2019-08-16 11:41:47,737 ] [ DEBUG ] [ infer:131 ]  Op: NextIteration
[ 2019-08-16 11:41:47,737 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,738 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,738 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,738 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,738 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,738 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/sub/x
[ 2019-08-16 11:41:47,738 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,738 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,738 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,738 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1.0
[ 2019-08-16 11:41:47,738 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,738 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/concat_1/axis
[ 2019-08-16 11:41:47,738 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,738 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,738 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,738 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,738 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,738 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/split/split_dim
[ 2019-08-16 11:41:47,738 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,739 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,739 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,739 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,739 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,739 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/concat/axis
[ 2019-08-16 11:41:47,739 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,739 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,739 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,739 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,739 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,739 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/add/y
[ 2019-08-16 11:41:47,739 ] [ DEBUG ] [ infer:131 ]  Op: Const
[ 2019-08-16 11:41:47,739 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,739 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,739 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,739 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,739 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/add
[ 2019-08-16 11:41:47,739 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,740 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,740 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,740 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 1
[ 2019-08-16 11:41:47,740 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,740 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,740 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,740 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/NextIteration
[ 2019-08-16 11:41:47,740 ] [ DEBUG ] [ infer:131 ]  Op: NextIteration
[ 2019-08-16 11:41:47,740 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,740 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,740 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,740 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,740 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,740 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArray_1
[ 2019-08-16 11:41:47,740 ] [ DEBUG ] [ infer:131 ]  Op: TensorArrayV3
[ 2019-08-16 11:41:47,741 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,741 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,741 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,741 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArray_1/Output_0/Data_
[ 2019-08-16 11:41:47,741 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,741 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,741 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/TensorArrayReadV3/Enter
[ 2019-08-16 11:41:47,741 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,741 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,741 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArray_1/Output_0/Data_
[ 2019-08-16 11:41:47,741 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,741 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArray_1/Output_0/Data_
[ 2019-08-16 11:41:47,741 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,741 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArrayUnstack/TensorArrayScatter/TensorArrayScatterV3
[ 2019-08-16 11:41:47,741 ] [ DEBUG ] [ infer:131 ]  Op: TensorArrayScatterV3
[ 2019-08-16 11:41:47,742 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,742 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArray_1/Output_0/Data_
[ 2019-08-16 11:41:47,742 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [28], value = [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
 24 25 26 27]
[ 2019-08-16 11:41:47,742 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [ 28   1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,742 ] [ DEBUG ] [ infer:32 ]  input[3]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,742 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,742 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,742 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,742 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/TensorArrayReadV3/Enter_1
[ 2019-08-16 11:41:47,742 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,742 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,742 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,742 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,742 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,743 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,743 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/TensorArrayReadV3
[ 2019-08-16 11:41:47,743 ] [ DEBUG ] [ infer:131 ]  Op: TensorArrayReadV3
[ 2019-08-16 11:41:47,743 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,743 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArray_1/Output_0/Data_
[ 2019-08-16 11:41:47,743 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,743 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,743 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,743 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,743 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,743 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArray
[ 2019-08-16 11:41:47,743 ] [ DEBUG ] [ infer:131 ]  Op: TensorArrayV3
[ 2019-08-16 11:41:47,743 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,743 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,743 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,744 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArray/Output_0/Data_
[ 2019-08-16 11:41:47,744 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,744 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,744 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/TensorArrayWrite/TensorArrayWriteV3/Enter
[ 2019-08-16 11:41:47,744 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,744 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,744 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArray/Output_0/Data_
[ 2019-08-16 11:41:47,744 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,744 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArray/Output_0/Data_
[ 2019-08-16 11:41:47,744 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,744 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Enter_2
[ 2019-08-16 11:41:47,744 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,744 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,744 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,744 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,744 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,744 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,745 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Merge_2
[ 2019-08-16 11:41:47,745 ] [ DEBUG ] [ infer:131 ]  Op: Merge
[ 2019-08-16 11:41:47,745 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,745 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,745 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = <UNKNOWN>, value = <UNKNOWN>
[ 2019-08-16 11:41:47,745 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,745 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,745 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,745 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Switch_2
[ 2019-08-16 11:41:47,745 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,745 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,745 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,745 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,745 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,745 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,745 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,746 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,746 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Identity_2
[ 2019-08-16 11:41:47,746 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,746 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,746 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,746 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,746 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,746 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,746 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/Shape
[ 2019-08-16 11:41:47,746 ] [ DEBUG ] [ infer:131 ]  Op: ShapeOf
[ 2019-08-16 11:41:47,746 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,746 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 28   1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,746 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,747 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [3], value = [ 28   1 192]
[ 2019-08-16 11:41:47,747 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,747 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/strided_slice
[ 2019-08-16 11:41:47,747 ] [ DEBUG ] [ infer:131 ]  Op: StridedSlice
[ 2019-08-16 11:41:47,747 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,748 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [3], value = [ 28   1 192]
[ 2019-08-16 11:41:47,748 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,748 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [1], value = [2]
[ 2019-08-16 11:41:47,748 ] [ DEBUG ] [ infer:32 ]  input[3]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,748 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,748 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,748 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,748 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/GRUCellZeroState/ExpandDims
[ 2019-08-16 11:41:47,748 ] [ DEBUG ] [ infer:131 ]  Op: Unsqueeze
[ 2019-08-16 11:41:47,749 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,749 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 1
[ 2019-08-16 11:41:47,749 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [0]
[ 2019-08-16 11:41:47,749 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,749 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,749 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,749 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/GRUCellZeroState/concat
[ 2019-08-16 11:41:47,749 ] [ DEBUG ] [ infer:131 ]  Op: ConcatV2
[ 2019-08-16 11:41:47,749 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,750 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = [1]
[ 2019-08-16 11:41:47,750 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [1], value = [96]
[ 2019-08-16 11:41:47,750 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,750 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [2], value = [ 1 96]
[ 2019-08-16 11:41:47,750 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,750 ] [ DEBUG ] [ infer:130 ]  Partial infer for 771
[ 2019-08-16 11:41:47,750 ] [ DEBUG ] [ infer:131 ]  Op: Broadcast
[ 2019-08-16 11:41:47,750 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,750 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0.0
[ 2019-08-16 11:41:47,750 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [2], value = [ 1 96]
[ 2019-08-16 11:41:47,750 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,751 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = [[0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
  0. 0. 0. 0. 0. 0. 0. ...
[ 2019-08-16 11:41:47,751 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,751 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Enter_3
[ 2019-08-16 11:41:47,751 ] [ DEBUG ] [ infer:131 ]  Op: Enter
[ 2019-08-16 11:41:47,751 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,752 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = [[0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
  0. 0. 0. 0. 0. 0. 0. ...
[ 2019-08-16 11:41:47,752 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,753 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = [[0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
  0. 0. 0. 0. 0. 0. 0. ...
[ 2019-08-16 11:41:47,753 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,753 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Merge_3
[ 2019-08-16 11:41:47,753 ] [ DEBUG ] [ infer:131 ]  Op: Merge
[ 2019-08-16 11:41:47,753 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,754 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = [[0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
  0. 0. 0. 0. 0. 0. 0. ...
[ 2019-08-16 11:41:47,754 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = <UNKNOWN>, value = <UNKNOWN>
[ 2019-08-16 11:41:47,754 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,754 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,754 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,754 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Switch_3
[ 2019-08-16 11:41:47,754 ] [ DEBUG ] [ infer:131 ]  Op: Switch
[ 2019-08-16 11:41:47,754 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,755 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,755 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,755 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,755 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,755 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,755 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Identity_3
[ 2019-08-16 11:41:47,755 ] [ DEBUG ] [ infer:131 ]  Op: Identity
[ 2019-08-16 11:41:47,755 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,755 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,755 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,755 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,755 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,755 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/concat
[ 2019-08-16 11:41:47,755 ] [ DEBUG ] [ infer:131 ]  Op: ConcatV2
[ 2019-08-16 11:41:47,756 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,756 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,756 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,756 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,756 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 288], value = <UNKNOWN>
[ 2019-08-16 11:41:47,756 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,756 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/MatMul
[ 2019-08-16 11:41:47,756 ] [ DEBUG ] [ infer:131 ]  Op: MatMul
[ 2019-08-16 11:41:47,756 ] [ DEBUG ] [ matmul:30 ]  matmul shapes: [array([  1, 288]), array([288, 192])]
[ 2019-08-16 11:41:47,756 ] [ DEBUG ] [ matmul:63 ]  shape_tuple: (array([1]), array([192]))
[ 2019-08-16 11:41:47,757 ] [ DEBUG ] [ matmul:66 ]  matmul shape: [  1 192]
[ 2019-08-16 11:41:47,757 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,757 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 288], value = <UNKNOWN>
[ 2019-08-16 11:41:47,757 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [288 192], value = [[-0.02369513 -0.41043222 -0.1847231  ...  0.17719221 -0.09955259
  -0.23327482]
 [-0.17745     0...
[ 2019-08-16 11:41:47,757 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,757 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,757 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,757 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/BiasAdd
[ 2019-08-16 11:41:47,758 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,758 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,758 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,759 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [192], value = [0.8774892  0.9430615  0.661636   0.93019795 0.9606291  0.9283228
 0.99488705 0.8112845  0.851202...
[ 2019-08-16 11:41:47,759 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,759 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,759 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,759 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/Sigmoid
[ 2019-08-16 11:41:47,759 ] [ DEBUG ] [ infer:131 ]  Op: Sigmoid
[ 2019-08-16 11:41:47,759 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,759 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,759 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,760 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,760 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,760 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/split
[ 2019-08-16 11:41:47,760 ] [ DEBUG ] [ infer:131 ]  Op: Split
[ 2019-08-16 11:41:47,760 ] [ DEBUG ] [ split:140 ]  input shape for split: [  1 192], should be split along 1 dim
[ 2019-08-16 11:41:47,760 ] [ DEBUG ] [ split:142 ]  split_dim_size type = <class 'numpy.int64'>
[ 2019-08-16 11:41:47,760 ] [ DEBUG ] [ split:150 ]  split_dim_size = 192, node.num_split = 2, div = 96.0, typeof div = <class 'numpy.float64'>
[ WARNING ]  Removed "axis" attribute from the scope of the model relayout pass because len(input.shape) == 2 != 4 for node rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/split
[ 2019-08-16 11:41:47,760 ] [ DEBUG ] [ split:118 ]  output shapes after split: [array([ 1, 96]), array([ 1, 96])]
[ 2019-08-16 11:41:47,760 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,760 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,760 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,761 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,761 ] [ DEBUG ] [ infer:32 ]  output[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,761 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,761 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/sub/negate_
[ 2019-08-16 11:41:47,761 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,761 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,761 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,761 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = -1
[ 2019-08-16 11:41:47,761 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,761 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,761 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,761 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/sub/add_
[ 2019-08-16 11:41:47,761 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,761 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,762 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,762 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 1.0
[ 2019-08-16 11:41:47,762 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,762 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,762 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,762 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/mul_1
[ 2019-08-16 11:41:47,762 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,762 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,762 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,762 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,762 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,762 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,762 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,763 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/mul
[ 2019-08-16 11:41:47,763 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,763 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,763 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,763 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,763 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,763 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,763 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,763 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/concat_1
[ 2019-08-16 11:41:47,763 ] [ DEBUG ] [ infer:131 ]  Op: ConcatV2
[ 2019-08-16 11:41:47,763 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,763 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,764 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,764 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,764 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1 288], value = <UNKNOWN>
[ 2019-08-16 11:41:47,764 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,764 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/MatMul_1
[ 2019-08-16 11:41:47,764 ] [ DEBUG ] [ infer:131 ]  Op: MatMul
[ 2019-08-16 11:41:47,764 ] [ DEBUG ] [ matmul:30 ]  matmul shapes: [array([  1, 288]), array([288,  96])]
[ 2019-08-16 11:41:47,764 ] [ DEBUG ] [ matmul:63 ]  shape_tuple: (array([1]), array([96]))
[ 2019-08-16 11:41:47,764 ] [ DEBUG ] [ matmul:66 ]  matmul shape: [ 1 96]
[ 2019-08-16 11:41:47,764 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,764 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1 288], value = <UNKNOWN>
[ 2019-08-16 11:41:47,765 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [288  96], value = [[-0.08617257  0.00440311  0.05893075 ...  0.07117532 -0.2046704
  -0.2270496 ]
 [ 0.08344617  0....
[ 2019-08-16 11:41:47,765 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,765 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,765 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,765 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/BiasAdd_1
[ 2019-08-16 11:41:47,765 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,765 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,765 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,766 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [96], value = [ 0.14700736 -0.04902952 -0.26648712 -0.27668694  0.1791299  -0.20344824
  0.2814196  -0.05231852...
[ 2019-08-16 11:41:47,766 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,766 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,766 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,766 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/Tanh
[ 2019-08-16 11:41:47,766 ] [ DEBUG ] [ infer:131 ]  Op: Tanh
[ 2019-08-16 11:41:47,766 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,767 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,767 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,767 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,767 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,767 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/mul_2
[ 2019-08-16 11:41:47,767 ] [ DEBUG ] [ infer:131 ]  Op: Mul
[ 2019-08-16 11:41:47,767 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,767 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,767 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,767 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,767 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,767 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,767 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/add
[ 2019-08-16 11:41:47,767 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,768 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,768 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,768 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,768 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,768 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,768 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,768 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/NextIteration_3
[ 2019-08-16 11:41:47,768 ] [ DEBUG ] [ infer:131 ]  Op: NextIteration
[ 2019-08-16 11:41:47,768 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,768 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,768 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,768 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,768 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,768 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/TensorArrayWrite/TensorArrayWriteV3
[ 2019-08-16 11:41:47,769 ] [ DEBUG ] [ infer:131 ]  Op: TensorArrayWriteV3
[ 2019-08-16 11:41:47,769 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,769 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArray/Output_0/Data_
[ 2019-08-16 11:41:47,769 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,769 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [ 1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,769 ] [ DEBUG ] [ infer:32 ]  input[3]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,769 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,769 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,769 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,769 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/NextIteration_2
[ 2019-08-16 11:41:47,769 ] [ DEBUG ] [ infer:131 ]  Op: NextIteration
[ 2019-08-16 11:41:47,769 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,769 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,769 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,769 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,770 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,770 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Exit_2
[ 2019-08-16 11:41:47,770 ] [ DEBUG ] [ infer:131 ]  Op: Exit
[ 2019-08-16 11:41:47,770 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,770 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,770 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,770 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,770 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,770 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArrayStack/TensorArraySizeV3
[ 2019-08-16 11:41:47,770 ] [ DEBUG ] [ infer:131 ]  Op: TensorArraySizeV3
[ 2019-08-16 11:41:47,770 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,770 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArray/Output_0/Data_
[ 2019-08-16 11:41:47,770 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,770 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,770 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [], value = 28
[ 2019-08-16 11:41:47,770 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,770 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArrayStack/range
[ 2019-08-16 11:41:47,771 ] [ DEBUG ] [ infer:131 ]  Op: Range
[ 2019-08-16 11:41:47,771 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,771 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = 0
[ 2019-08-16 11:41:47,771 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 28
[ 2019-08-16 11:41:47,771 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [], value = 1
[ 2019-08-16 11:41:47,771 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,771 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [28], value = [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
 24 25 26 27]
[ 2019-08-16 11:41:47,771 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,771 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArrayStack/TensorArrayGatherV3
[ 2019-08-16 11:41:47,771 ] [ DEBUG ] [ infer:131 ]  Op: TensorArrayGatherV3
[ 2019-08-16 11:41:47,771 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,771 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [], value = rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArray/Output_0/Data_
[ 2019-08-16 11:41:47,772 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [28], value = [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
 24 25 26 27]
[ 2019-08-16 11:41:47,772 ] [ DEBUG ] [ infer:32 ]  input[2]: shape = [], value = <UNKNOWN>
[ 2019-08-16 11:41:47,772 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,772 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [28  1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,772 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,772 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/transpose_1
[ 2019-08-16 11:41:47,772 ] [ DEBUG ] [ infer:131 ]  Op: Transpose
[ 2019-08-16 11:41:47,772 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,773 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [28  1 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,773 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [3], value = [1 0 2]
[ 2019-08-16 11:41:47,773 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,773 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 28 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,773 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,773 ] [ DEBUG ] [ infer:130 ]  Partial infer for rnn/stack_bidirectional_rnn/cell_1/concat
[ 2019-08-16 11:41:47,773 ] [ DEBUG ] [ infer:131 ]  Op: ConcatV2
[ 2019-08-16 11:41:47,773 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,773 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 28 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,773 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [ 1 28 96], value = <UNKNOWN>
[ 2019-08-16 11:41:47,773 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,774 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1  28 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,774 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,774 ] [ DEBUG ] [ infer:130 ]  Partial infer for Reshape
[ 2019-08-16 11:41:47,774 ] [ DEBUG ] [ infer:131 ]  Op: Reshape
[ 2019-08-16 11:41:47,774 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,774 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1  28 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,774 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [4], value = [ -1  28   1 192]
[ 2019-08-16 11:41:47,774 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,774 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [  1  28   1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,774 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,774 ] [ DEBUG ] [ infer:130 ]  Partial infer for conv2d/Conv2D
[ 2019-08-16 11:41:47,775 ] [ DEBUG ] [ infer:131 ]  Op: Conv2D
[ 2019-08-16 11:41:47,775 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,775 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [  1  28   1 192], value = <UNKNOWN>
[ 2019-08-16 11:41:47,776 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [  1   1 192  41], value = [[[[ 0.10394846  0.266858   -0.28928807 ...  0.04898277 -0.1620783
    -0.08559268]
   [ 0.064426...
[ 2019-08-16 11:41:47,776 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,776 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 28  1 41], value = <UNKNOWN>
[ 2019-08-16 11:41:47,776 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,776 ] [ DEBUG ] [ infer:130 ]  Partial infer for conv2d/BiasAdd
[ 2019-08-16 11:41:47,776 ] [ DEBUG ] [ infer:131 ]  Op: Add
[ 2019-08-16 11:41:47,776 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,776 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 28  1 41], value = <UNKNOWN>
[ 2019-08-16 11:41:47,776 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [41], value = [ 3.97828430e-01 -4.31426242e-03 -3.57889608e-02 -3.92717533e-02
 -8.06604549e-02 -8.69704634e-02...
[ 2019-08-16 11:41:47,776 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,777 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 28  1 41], value = <UNKNOWN>
[ 2019-08-16 11:41:47,777 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,777 ] [ DEBUG ] [ infer:130 ]  Partial infer for Reshape_1
[ 2019-08-16 11:41:47,777 ] [ DEBUG ] [ infer:131 ]  Op: Reshape
[ 2019-08-16 11:41:47,777 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:47,777 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 28  1 41], value = <UNKNOWN>
[ 2019-08-16 11:41:47,777 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [3], value = [-1 28 41]
[ 2019-08-16 11:41:47,777 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:47,777 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 28 41], value = <UNKNOWN>
[ 2019-08-16 11:41:47,778 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:47,778 ] [ DEBUG ] [ infer:130 ]  Partial infer for Abs
[ 2019-08-16 11:41:47,778 ] [ DEBUG ] [ infer:131 ]  Op: Abs
[ INFO ]  Called "tf_native_tf_node_infer" for node "Abs"
[ 2019-08-16 11:41:47,779 ] [ DEBUG ] [ tf:222 ]  Added placeholder with name 'Reshape_1_port_0_ie_placeholder'
[ 2019-08-16 11:41:47,779 ] [ DEBUG ] [ tf:235 ]  update_input_in_pbs: replace input 'Reshape_1' with input 'Reshape_1_port_0_ie_placeholder'
[ 2019-08-16 11:41:47,779 ] [ DEBUG ] [ tf:243 ]  Replacing input '0' of the node 'Abs' with placeholder 'Reshape_1_port_0_ie_placeholder'
[ 2019-08-16 11:41:48,270 ] [ DEBUG ] [ tf:142 ]  Inferred shape of the output tensor with index '0' of the node 'Abs': '[ 1 28 41]'
[ 2019-08-16 11:41:48,271 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:48,271 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 28 41], value = <UNKNOWN>
[ 2019-08-16 11:41:48,271 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:48,271 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 28 41], value = <UNKNOWN>
[ 2019-08-16 11:41:48,271 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:48,271 ] [ DEBUG ] [ infer:130 ]  Partial infer for Max
[ 2019-08-16 11:41:48,271 ] [ DEBUG ] [ infer:131 ]  Op: ReduceMax
[ 2019-08-16 11:41:48,271 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:48,272 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 28 41], value = <UNKNOWN>
[ 2019-08-16 11:41:48,272 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 2
[ 2019-08-16 11:41:48,272 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:48,272 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 28], value = <UNKNOWN>
[ 2019-08-16 11:41:48,272 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:48,272 ] [ DEBUG ] [ infer:130 ]  Partial infer for Sign
[ 2019-08-16 11:41:48,272 ] [ DEBUG ] [ infer:131 ]  Op: Sign
[ INFO ]  Called "tf_native_tf_node_infer" for node "Sign"
[ 2019-08-16 11:41:48,273 ] [ DEBUG ] [ tf:222 ]  Added placeholder with name 'Max_port_0_ie_placeholder'
[ 2019-08-16 11:41:48,273 ] [ DEBUG ] [ tf:235 ]  update_input_in_pbs: replace input 'Max' with input 'Max_port_0_ie_placeholder'
[ 2019-08-16 11:41:48,274 ] [ DEBUG ] [ tf:243 ]  Replacing input '0' of the node 'Sign' with placeholder 'Max_port_0_ie_placeholder'
[ 2019-08-16 11:41:48,277 ] [ DEBUG ] [ tf:142 ]  Inferred shape of the output tensor with index '0' of the node 'Sign': '[ 1 28]'
[ 2019-08-16 11:41:48,277 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:48,277 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 28], value = <UNKNOWN>
[ 2019-08-16 11:41:48,277 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:48,277 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [ 1 28], value = <UNKNOWN>
[ 2019-08-16 11:41:48,277 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:48,277 ] [ DEBUG ] [ infer:130 ]  Partial infer for Sum
[ 2019-08-16 11:41:48,277 ] [ DEBUG ] [ infer:131 ]  Op: ReduceSum
[ 2019-08-16 11:41:48,278 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:48,278 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [ 1 28], value = <UNKNOWN>
[ 2019-08-16 11:41:48,278 ] [ DEBUG ] [ infer:32 ]  input[1]: shape = [], value = 1
[ 2019-08-16 11:41:48,278 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:48,278 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:48,278 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:48,278 ] [ DEBUG ] [ infer:130 ]  Partial infer for seq_len
[ 2019-08-16 11:41:48,278 ] [ DEBUG ] [ infer:131 ]  Op: Cast
[ 2019-08-16 11:41:48,278 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:48,278 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:48,278 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:48,278 ] [ DEBUG ] [ infer:32 ]  output[0]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:48,279 ] [ DEBUG ] [ infer:129 ]  --------------------
[ 2019-08-16 11:41:48,279 ] [ DEBUG ] [ infer:130 ]  Partial infer for seq_len/sink_port_0
[ 2019-08-16 11:41:48,279 ] [ DEBUG ] [ infer:131 ]  Op: Result
[ 2019-08-16 11:41:48,279 ] [ DEBUG ] [ infer:142 ]  Inputs:
[ 2019-08-16 11:41:48,279 ] [ DEBUG ] [ infer:32 ]  input[0]: shape = [1], value = <UNKNOWN>
[ 2019-08-16 11:41:48,279 ] [ DEBUG ] [ infer:144 ]  Outputs:
[ 2019-08-16 11:41:48,297 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.SwitchMergeOptimization.SwitchMergeMiddleReplacer'>
[ 2019-08-16 11:41:48,339 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.ReverseV2ToReverseSequence.ReverseToReverseSequence'>
[ 2019-08-16 11:41:48,390 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.DeleteControlFlowEdges.DeleteControlFlowEdges'>
[ 2019-08-16 11:41:48,392 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from batch_normalization/cond_2/switch_f/Output_0/Data_ to batch_normalization/cond_2/ones_like/Shape
[ 2019-08-16 11:41:48,392 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from batch_normalization/cond_2/switch_f/Output_0/Data_ to batch_normalization/cond_2/ones_like/Const
[ 2019-08-16 11:41:48,392 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from batch_normalization/cond_3/switch_f/Output_0/Data_ to batch_normalization/cond_3/zeros_like
[ 2019-08-16 11:41:48,392 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from batch_normalization_1/cond/switch_t/Output_0/Data_ to batch_normalization_1/cond/Const
[ 2019-08-16 11:41:48,392 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from batch_normalization_1/cond/switch_t/Output_0/Data_ to batch_normalization_1/cond/Const_1
[ 2019-08-16 11:41:48,392 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from batch_normalization_2/cond/switch_t/Output_0/Data_ to batch_normalization_2/cond/Const
[ 2019-08-16 11:41:48,392 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from batch_normalization_2/cond/switch_t/Output_0/Data_ to batch_normalization_2/cond/Const_1
[ 2019-08-16 11:41:48,392 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Identity/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/add/y
[ 2019-08-16 11:41:48,392 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Identity/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/concat/axis
[ 2019-08-16 11:41:48,392 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Identity/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/split/split_dim
[ 2019-08-16 11:41:48,392 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Identity/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/concat_1/axis
[ 2019-08-16 11:41:48,392 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Identity/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/sub/x
[ 2019-08-16 11:41:48,392 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Identity/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/add_1/y
[ 2019-08-16 11:41:48,392 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Identity/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/add/y
[ 2019-08-16 11:41:48,392 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Identity/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/concat/axis
[ 2019-08-16 11:41:48,392 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Identity/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/split/split_dim
[ 2019-08-16 11:41:48,392 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Identity/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/concat_1/axis
[ 2019-08-16 11:41:48,392 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Identity/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/sub/x
[ 2019-08-16 11:41:48,392 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Identity/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/add_1/y
[ 2019-08-16 11:41:48,393 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Identity/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/add/y
[ 2019-08-16 11:41:48,393 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Identity/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/concat/axis
[ 2019-08-16 11:41:48,393 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Identity/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/split/split_dim
[ 2019-08-16 11:41:48,393 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Identity/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/concat_1/axis
[ 2019-08-16 11:41:48,393 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Identity/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/sub/x
[ 2019-08-16 11:41:48,393 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Identity/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/add_1/y
[ 2019-08-16 11:41:48,393 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Identity/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/add/y
[ 2019-08-16 11:41:48,393 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Identity/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/concat/axis
[ 2019-08-16 11:41:48,393 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Identity/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/split/split_dim
[ 2019-08-16 11:41:48,393 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Identity/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/concat_1/axis
[ 2019-08-16 11:41:48,393 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Identity/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/sub/x
[ 2019-08-16 11:41:48,393 ] [ DEBUG ] [ DeleteControlFlowEdges:37 ]  Removing control flow edge from rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Identity/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/add_1/y
[ 2019-08-16 11:41:48,409 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.quantize_fuses.MarkNodesToFuseUpToFakeQuantize'>
[ 2019-08-16 11:41:48,433 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.quantize_fuses.FakeQuantizeFuse'>
[ 2019-08-16 11:41:48,449 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.BinarizeWeightsM1P1.BinarizeWeightsM1P1'>
[ 2019-08-16 11:41:48,492 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.ReluQuantizeFuse.ReluFakeQuantizeMark'>
[ 2019-08-16 11:41:48,534 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.ReluQuantizeFuse.ReluQuantizeFuse'>
[ 2019-08-16 11:41:48,577 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.ReluQuantizeFuse.ClampQuantizeMark'>
[ 2019-08-16 11:41:48,620 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.AddIsCyclicAttribute.AddIsCyclicAttribute'>
[ 2019-08-16 11:41:48,641 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.TensorIteratorInput.SmartInputMatcher'>
[ 2019-08-16 11:41:48,677 ] [ DEBUG ] [ TensorIteratorInput:129 ]  ================== SmartInputFind ===============
[ 2019-08-16 11:41:48,678 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'op': 'TensorIteratorInput', 'axis': 0, 'start': None, 'end': None, 'stride': None, 'part_size': None, 'in_ports_count': 3, 'out_ports_count': 1, 'infer': <function TensorIteratorInput.input_infer at 0x7fe50bcf7158>, 'external_port_id': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArray_1/Output_0/Data_', 'internal_layer_id': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/TensorArrayReadV3/Output_0/Data_', 'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/TensorArrayReadV3/TensorIteratorInput_', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a270510>), 'name', 'precision', 'type'], [('data', ['external_port_id', 'internal_layer_id', 'internal_port_id', 'axis', 'start', 'stride', 'part_size'], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}, 2: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:48,678 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/TensorArrayReadV3/Output_0/Data_', 'kind': 'data', 'shape': array([  1, 192]), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/TensorArrayReadV3', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True}
[ 2019-08-16 11:41:48,678 ] [ DEBUG ] [ TensorIteratorInput:129 ]  ================== SmartInputFind ===============
[ 2019-08-16 11:41:48,679 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'op': 'TensorIteratorInput', 'axis': 0, 'start': None, 'end': None, 'stride': None, 'part_size': None, 'in_ports_count': 3, 'out_ports_count': 1, 'infer': <function TensorIteratorInput.input_infer at 0x7fe50bcf7158>, 'external_port_id': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArray_1/Output_0/Data_', 'internal_layer_id': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/TensorArrayReadV3/Output_0/Data_', 'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/TensorArrayReadV3/TensorIteratorInput_', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a303ea0>), 'name', 'precision', 'type'], [('data', ['external_port_id', 'internal_layer_id', 'internal_port_id', 'axis', 'start', 'stride', 'part_size'], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}, 2: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:48,679 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/TensorArrayReadV3/Output_0/Data_', 'kind': 'data', 'shape': array([  1, 192]), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/TensorArrayReadV3', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True}
[ 2019-08-16 11:41:48,679 ] [ DEBUG ] [ TensorIteratorInput:129 ]  ================== SmartInputFind ===============
[ 2019-08-16 11:41:48,679 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'op': 'TensorIteratorInput', 'axis': 0, 'start': None, 'end': None, 'stride': None, 'part_size': None, 'in_ports_count': 3, 'out_ports_count': 1, 'infer': <function TensorIteratorInput.input_infer at 0x7fe50bcf7158>, 'external_port_id': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArray_1/Output_0/Data_', 'internal_layer_id': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/TensorArrayReadV3/Output_0/Data_', 'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/TensorArrayReadV3/TensorIteratorInput_', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a371ae8>), 'name', 'precision', 'type'], [('data', ['external_port_id', 'internal_layer_id', 'internal_port_id', 'axis', 'start', 'stride', 'part_size'], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}, 2: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:48,680 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/TensorArrayReadV3/Output_0/Data_', 'kind': 'data', 'shape': array([ 1, 96]), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/TensorArrayReadV3', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True}
[ 2019-08-16 11:41:48,680 ] [ DEBUG ] [ TensorIteratorInput:129 ]  ================== SmartInputFind ===============
[ 2019-08-16 11:41:48,680 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'op': 'TensorIteratorInput', 'axis': 0, 'start': None, 'end': None, 'stride': None, 'part_size': None, 'in_ports_count': 3, 'out_ports_count': 1, 'infer': <function TensorIteratorInput.input_infer at 0x7fe50bcf7158>, 'external_port_id': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArray_1/Output_0/Data_', 'internal_layer_id': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/TensorArrayReadV3/Output_0/Data_', 'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/TensorArrayReadV3/TensorIteratorInput_', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a495e18>), 'name', 'precision', 'type'], [('data', ['external_port_id', 'internal_layer_id', 'internal_port_id', 'axis', 'start', 'stride', 'part_size'], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}, 2: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:48,680 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/TensorArrayReadV3/Output_0/Data_', 'kind': 'data', 'shape': array([ 1, 96]), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/TensorArrayReadV3', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True}
[ 2019-08-16 11:41:48,694 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.TensorIteratorOutput.SmartOutputMatcher'>
[ 2019-08-16 11:41:48,730 ] [ DEBUG ] [ TensorIteratorOutput:133 ]  ================== SmartOutputFind ===============
[ 2019-08-16 11:41:48,731 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'op': 'TensorIteratorOutput', 'axis': 0, 'start': None, 'end': None, 'stride': None, 'part_size': None, 'in_ports_count': 3, 'out_ports_count': 1, 'infer': <function TensorIteratorOutput.input_infer at 0x7fe50bcf72f0>, 'external_port_id': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArray/Output_0/Data_', 'internal_layer_id': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_', 'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/TensorArrayWrite/TensorArrayWriteV3/TensorIteratorOutput_', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a7d8268>), 'name', 'precision', 'type'], [('data', ['external_port_id', 'internal_layer_id', 'internal_port_id', 'axis', 'start', 'stride', 'part_size'], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}, 2: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:48,731 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArrayStack/TensorArrayGatherV3/Output_0/Data_', 'kind': 'data', 'shape': array([28,  1, 96]), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArrayStack/TensorArrayGatherV3', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True}
[ 2019-08-16 11:41:48,731 ] [ DEBUG ] [ TensorIteratorOutput:133 ]  ================== SmartOutputFind ===============
[ 2019-08-16 11:41:48,732 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'op': 'TensorIteratorOutput', 'axis': 0, 'start': None, 'end': None, 'stride': None, 'part_size': None, 'in_ports_count': 3, 'out_ports_count': 1, 'infer': <function TensorIteratorOutput.input_infer at 0x7fe50bcf72f0>, 'external_port_id': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArray/Output_0/Data_', 'internal_layer_id': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_', 'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/TensorArrayWrite/TensorArrayWriteV3/TensorIteratorOutput_', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a42f400>), 'name', 'precision', 'type'], [('data', ['external_port_id', 'internal_layer_id', 'internal_port_id', 'axis', 'start', 'stride', 'part_size'], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}, 2: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:48,732 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArrayStack/TensorArrayGatherV3/Output_0/Data_', 'kind': 'data', 'shape': array([28,  1, 96]), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArrayStack/TensorArrayGatherV3', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True}
[ 2019-08-16 11:41:48,732 ] [ DEBUG ] [ TensorIteratorOutput:133 ]  ================== SmartOutputFind ===============
[ 2019-08-16 11:41:48,732 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'op': 'TensorIteratorOutput', 'axis': 0, 'start': None, 'end': None, 'stride': None, 'part_size': None, 'in_ports_count': 3, 'out_ports_count': 1, 'infer': <function TensorIteratorOutput.input_infer at 0x7fe50bcf72f0>, 'external_port_id': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArray/Output_0/Data_', 'internal_layer_id': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_', 'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/TensorArrayWrite/TensorArrayWriteV3/TensorIteratorOutput_', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a32c950>), 'name', 'precision', 'type'], [('data', ['external_port_id', 'internal_layer_id', 'internal_port_id', 'axis', 'start', 'stride', 'part_size'], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}, 2: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:48,732 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArrayStack/TensorArrayGatherV3/Output_0/Data_', 'kind': 'data', 'shape': array([28,  1, 96]), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArrayStack/TensorArrayGatherV3', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True}
[ 2019-08-16 11:41:48,732 ] [ DEBUG ] [ TensorIteratorOutput:133 ]  ================== SmartOutputFind ===============
[ 2019-08-16 11:41:48,733 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'op': 'TensorIteratorOutput', 'axis': 0, 'start': None, 'end': None, 'stride': None, 'part_size': None, 'in_ports_count': 3, 'out_ports_count': 1, 'infer': <function TensorIteratorOutput.input_infer at 0x7fe50bcf72f0>, 'external_port_id': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArray/Output_0/Data_', 'internal_layer_id': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_', 'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/TensorArrayWrite/TensorArrayWriteV3/TensorIteratorOutput_', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a329840>), 'name', 'precision', 'type'], [('data', ['external_port_id', 'internal_layer_id', 'internal_port_id', 'axis', 'start', 'stride', 'part_size'], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}, 2: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:48,733 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArrayStack/TensorArrayGatherV3/Output_0/Data_', 'kind': 'data', 'shape': array([28,  1, 96]), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArrayStack/TensorArrayGatherV3', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True}
[ 2019-08-16 11:41:48,746 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.TensorIteratorOutput.SimpleOutputMatcher'>
[ 2019-08-16 11:41:48,780 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.TensorIteratorCondition.LoopConditionMatcher'>
[ 2019-08-16 11:41:48,780 ] [ DEBUG ] [ TensorIteratorCondition:84 ]  +++++++++++++++ ConditionMatching ++++++++++++++++
[ 2019-08-16 11:41:48,841 ] [ DEBUG ] [ TensorIteratorCondition:224 ]  ================== ConditionFind ===============
[ 2019-08-16 11:41:48,842 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'op': 'TensorIteratorCondition', 'in_ports_count': 2, 'out_ports_count': 2, 'infer': <function TensorIteratorCondition.input_infer at 0x7fe50bcf7400>, 'time': {'init': 0, 'step': 1}, 'iter': {'init': 0, 'step': 1}, 'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/LoopCond/TensorIteratorCondition_', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a3bd9d8>), 'name', 'precision', 'type'], [('data', [], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}}, '_out_ports': {0: {}, 1: {}}}
[ 2019-08-16 11:41:48,842 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/LoopCond/Output_0/Data_', 'kind': 'data', 'shape': array([], dtype=int64), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/LoopCond', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True}
[ 2019-08-16 11:41:48,842 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Identity_1/Output_0/Data_', 'kind': 'data', 'shape': array([], dtype=int64), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Identity_1', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True}
[ 2019-08-16 11:41:48,842 ] [ DEBUG ] [ TensorIteratorCondition:224 ]  ================== ConditionFind ===============
[ 2019-08-16 11:41:48,843 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'op': 'TensorIteratorCondition', 'in_ports_count': 2, 'out_ports_count': 2, 'infer': <function TensorIteratorCondition.input_infer at 0x7fe50bcf7400>, 'time': {'init': 0, 'step': 1}, 'iter': {'init': 0, 'step': 1}, 'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/LoopCond/TensorIteratorCondition_', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a486a60>), 'name', 'precision', 'type'], [('data', [], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}}, '_out_ports': {0: {}, 1: {}}}
[ 2019-08-16 11:41:48,843 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/LoopCond/Output_0/Data_', 'kind': 'data', 'shape': array([], dtype=int64), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/LoopCond', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True}
[ 2019-08-16 11:41:48,843 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Identity_1/Output_0/Data_', 'kind': 'data', 'shape': array([], dtype=int64), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Identity_1', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True}
[ 2019-08-16 11:41:48,843 ] [ DEBUG ] [ TensorIteratorCondition:224 ]  ================== ConditionFind ===============
[ 2019-08-16 11:41:48,843 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'op': 'TensorIteratorCondition', 'in_ports_count': 2, 'out_ports_count': 2, 'infer': <function TensorIteratorCondition.input_infer at 0x7fe50bcf7400>, 'time': {'init': 0, 'step': 1}, 'iter': {'init': 0, 'step': 1}, 'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/LoopCond/TensorIteratorCondition_', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a410378>), 'name', 'precision', 'type'], [('data', [], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}}, '_out_ports': {0: {}, 1: {}}}
[ 2019-08-16 11:41:48,844 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/LoopCond/Output_0/Data_', 'kind': 'data', 'shape': array([], dtype=int64), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/LoopCond', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True}
[ 2019-08-16 11:41:48,844 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Identity_1/Output_0/Data_', 'kind': 'data', 'shape': array([], dtype=int64), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Identity_1', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True}
[ 2019-08-16 11:41:48,844 ] [ DEBUG ] [ TensorIteratorCondition:224 ]  ================== ConditionFind ===============
[ 2019-08-16 11:41:48,844 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'op': 'TensorIteratorCondition', 'in_ports_count': 2, 'out_ports_count': 2, 'infer': <function TensorIteratorCondition.input_infer at 0x7fe50bcf7400>, 'time': {'init': 0, 'step': 1}, 'iter': {'init': 0, 'step': 1}, 'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/LoopCond/TensorIteratorCondition_', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a36d730>), 'name', 'precision', 'type'], [('data', [], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}}, '_out_ports': {0: {}, 1: {}}}
[ 2019-08-16 11:41:48,844 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/LoopCond/Output_0/Data_', 'kind': 'data', 'shape': array([], dtype=int64), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/LoopCond', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True}
[ 2019-08-16 11:41:48,844 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Identity_1/Output_0/Data_', 'kind': 'data', 'shape': array([], dtype=int64), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Identity_1', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True}
[ 2019-08-16 11:41:48,856 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.TensorIteratorCondition.SimpleConditionMatcher'>
[ 2019-08-16 11:41:48,856 ] [ DEBUG ] [ TensorIteratorCondition:307 ]  +++++++++++++++ SimpleConditionMatching ++++++++++++++++
[ 2019-08-16 11:41:48,890 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.TensorIteratorCondition.DynamicDecoderConditionMatcher'>
[ 2019-08-16 11:41:48,890 ] [ DEBUG ] [ TensorIteratorCondition:415 ]  +++++++++++++++ DynamicDecoderConditionMatching ++++++++++++++++
[ 2019-08-16 11:41:48,921 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.TensorIteratorBackEdge.BackEdgesMatching'>
[ 2019-08-16 11:41:48,971 ] [ DEBUG ] [ TensorIteratorBackEdge:96 ]  ================== BackEdgeFind ===============
[ 2019-08-16 11:41:48,971 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'op': 'TensorIteratorBackEdge', 'in_ports_count': 3, 'out_ports_count': 1, 'infer': <function TensorIteratorBackEdge.input_infer at 0x7fe50bcf7510>, 'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Identity_3/TensorIteratorBackEdge_', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a364bf8>), 'name', 'precision', 'type'], [('data', [], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}, 2: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:48,972 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Identity_3/Output_0/Data_', 'kind': 'data', 'shape': array([ 1, 96]), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Identity_3', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True}
[ 2019-08-16 11:41:48,972 ] [ DEBUG ] [ TensorIteratorBackEdge:96 ]  ================== BackEdgeFind ===============
[ 2019-08-16 11:41:48,972 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'op': 'TensorIteratorBackEdge', 'in_ports_count': 3, 'out_ports_count': 1, 'infer': <function TensorIteratorBackEdge.input_infer at 0x7fe50bcf7510>, 'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Identity_3/TensorIteratorBackEdge_', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a2f1158>), 'name', 'precision', 'type'], [('data', [], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}, 2: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:48,972 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Identity_3/Output_0/Data_', 'kind': 'data', 'shape': array([ 1, 96]), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Identity_3', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True}
[ 2019-08-16 11:41:48,972 ] [ DEBUG ] [ TensorIteratorBackEdge:96 ]  ================== BackEdgeFind ===============
[ 2019-08-16 11:41:48,973 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'op': 'TensorIteratorBackEdge', 'in_ports_count': 3, 'out_ports_count': 1, 'infer': <function TensorIteratorBackEdge.input_infer at 0x7fe50bcf7510>, 'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Identity_3/TensorIteratorBackEdge_', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a2f1400>), 'name', 'precision', 'type'], [('data', [], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}, 2: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:48,973 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Identity_3/Output_0/Data_', 'kind': 'data', 'shape': array([ 1, 96]), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Identity_3', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True}
[ 2019-08-16 11:41:48,973 ] [ DEBUG ] [ TensorIteratorBackEdge:96 ]  ================== BackEdgeFind ===============
[ 2019-08-16 11:41:48,973 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'op': 'TensorIteratorBackEdge', 'in_ports_count': 3, 'out_ports_count': 1, 'infer': <function TensorIteratorBackEdge.input_infer at 0x7fe50bcf7510>, 'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Identity_3/TensorIteratorBackEdge_', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a2f16a8>), 'name', 'precision', 'type'], [('data', [], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}, 2: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:48,973 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Identity_3/Output_0/Data_', 'kind': 'data', 'shape': array([ 1, 96]), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Identity_3', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True}
[ 2019-08-16 11:41:48,984 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.TensorIteratorConditionChecker.ConditionChecks'>
[ 2019-08-16 11:41:48,984 ] [ DEBUG ] [ TensorIteratorConditionChecker:37 ]  +++++++++++++++ ConditionCheckerMatching ++++++++++++++++
[ 2019-08-16 11:41:49,016 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.DeleteNotExecutable.DeleteNotExecutable'>
[ 2019-08-16 11:41:49,016 ] [ DEBUG ] [ DeleteNotExecutable:41 ]  Removing the following not executable nodes: batch_normalization/cond/Switch_1
batch_normalization/cond/Switch_1/Output_1/Data_
batch_normalization/cond_1/Switch_1
batch_normalization/cond_1/Switch_1/Output_1/Data_
batch_normalization/cond_2/Switch_1
batch_normalization/cond_2/Switch_1/Output_1/Data_
batch_normalization/cond_3/Switch_1
batch_normalization/cond_3/Switch_1/Output_1/Data_
batch_normalization_1/cond/Const
batch_normalization_1/cond/Const/Output_0/Data_
batch_normalization_1/cond/Const_1
batch_normalization_1/cond/Const_1/Output_0/Data_
batch_normalization_1/cond/FusedBatchNorm
batch_normalization_1/cond/FusedBatchNorm/Output_0/Data_
batch_normalization_1/cond/FusedBatchNorm/Switch
batch_normalization_1/cond/FusedBatchNorm/Switch/Output_1/Data_
batch_normalization_1/cond/FusedBatchNorm/Switch_1
batch_normalization_1/cond/FusedBatchNorm/Switch_1/Output_1/Data_
batch_normalization_1/cond/FusedBatchNorm/Switch_2
batch_normalization_1/cond/FusedBatchNorm/Switch_2/Output_1/Data_
batch_normalization_1/cond/Switch
batch_normalization_1/cond/Switch/Output_1/Data_
batch_normalization_1/cond/switch_t
batch_normalization_1/cond/switch_t/Output_0/Data_
batch_normalization_2/cond/Const
batch_normalization_2/cond/Const/Output_0/Data_
batch_normalization_2/cond/Const_1
batch_normalization_2/cond/Const_1/Output_0/Data_
batch_normalization_2/cond/FusedBatchNorm
batch_normalization_2/cond/FusedBatchNorm/Output_0/Data_
batch_normalization_2/cond/FusedBatchNorm/Switch
batch_normalization_2/cond/FusedBatchNorm/Switch/Output_1/Data_
batch_normalization_2/cond/FusedBatchNorm/Switch_1
batch_normalization_2/cond/FusedBatchNorm/Switch_1/Output_1/Data_
batch_normalization_2/cond/FusedBatchNorm/Switch_2
batch_normalization_2/cond/FusedBatchNorm/Switch_2/Output_1/Data_
batch_normalization_2/cond/Switch
batch_normalization_2/cond/Switch/Output_1/Data_
batch_normalization_2/cond/switch_t
batch_normalization_2/cond/switch_t/Output_0/Data_
[ 2019-08-16 11:41:49,019 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
seq_len/sink_port_0
[ 2019-08-16 11:41:49,067 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 1514
1518
1522
1526
759
759/Output_0/Data_
761
766
766/Output_0/Data_
768
769
770
771
772
773
773/Output_0/Data_
776
776/Output_0/Data_
779
779/Output_0/Data_
782
785
788
788/Output_0/Data_
791
791/Output_0/Data_
794
797
800
Max/reduction_indices
PlaceholderWithDefault
PlaceholderWithDefault/input
Reshape/shape
Reshape_1/shape
Sum/reduction_indices
batch_normalization/Sqrt/pow_
batch_normalization/Sqrt/pow_/Output_0/Data_
batch_normalization/add
batch_normalization/add/Output_0/Data_
batch_normalization/add/y
batch_normalization/add/y/Output_0/Data_
batch_normalization/add_1
batch_normalization/add_1/Output_0/Data_
batch_normalization/add_2
batch_normalization/add_2/Output_0/Data_
batch_normalization/batchnorm/Rsqrt/pow_
batch_normalization/batchnorm/add
batch_normalization/batchnorm/add/y
batch_normalization/beta
batch_normalization/beta/read
batch_normalization/cond/Merge
batch_normalization/cond/Switch_2
batch_normalization/cond/pred_id
batch_normalization/cond_1/Merge
batch_normalization/cond_1/Switch_2
batch_normalization/cond_1/pred_id
batch_normalization/cond_2/Merge
batch_normalization/cond_2/Switch
batch_normalization/cond_2/Switch/Output_0/Data_
batch_normalization/cond_2/ones_like/Const
batch_normalization/cond_2/ones_like/Shape
batch_normalization/cond_2/pred_id
batch_normalization/cond_2/pred_id/Output_0/Data_
batch_normalization/cond_2/switch_f
batch_normalization/cond_2/switch_f/Output_0/Data_
batch_normalization/cond_3/Merge
batch_normalization/cond_3/Switch
batch_normalization/cond_3/Switch/Output_0/Data_
batch_normalization/cond_3/pred_id
batch_normalization/cond_3/pred_id/Output_0/Data_
batch_normalization/cond_3/switch_f
batch_normalization/cond_3/switch_f/Output_0/Data_
batch_normalization/cond_3/zeros_like
batch_normalization/gamma
batch_normalization/gamma/read
batch_normalization/moments/SquaredDifference/add_
batch_normalization/moments/SquaredDifference/add_/Output_0/Data_
batch_normalization/moments/SquaredDifference/negate_
batch_normalization/moments/SquaredDifference/negate_/Output_0/Data_
batch_normalization/moments/SquaredDifference/negate_const_
batch_normalization/moments/SquaredDifference/negate_const_/Output_0/Data_
batch_normalization/moments/SquaredDifference/squared_
batch_normalization/moments/SquaredDifference/squared_/Output_0/Data_
batch_normalization/moments/Squeeze
batch_normalization/moments/Squeeze/Dims
batch_normalization/moments/Squeeze/Dims/Output_0/Data_
batch_normalization/moments/Squeeze/Output_0/Data_
batch_normalization/moments/Squeeze_1
batch_normalization/moments/Squeeze_1/Dims
batch_normalization/moments/Squeeze_1/Dims/Output_0/Data_
batch_normalization/moments/Squeeze_1/Output_0/Data_
batch_normalization/moments/StopGradient
batch_normalization/moments/StopGradient/Output_0/Data_
batch_normalization/moments/mean
batch_normalization/moments/mean/Output_0/Data_
batch_normalization/moments/mean/reduction_indices
batch_normalization/moments/mean/reduction_indices/Output_0/Data_
batch_normalization/moments/variance
batch_normalization/moments/variance/Output_0/Data_
batch_normalization/moments/variance/reduction_indices
batch_normalization/moments/variance/reduction_indices/Output_0/Data_
batch_normalization/moving_mean
batch_normalization/moving_mean/read
batch_normalization/moving_variance
batch_normalization/moving_variance/read
batch_normalization/mul
batch_normalization/mul/Output_0/Data_
batch_normalization/mul_1
batch_normalization/mul_1/Output_0/Data_
batch_normalization/renorm_mean
batch_normalization/renorm_mean/Output_0/Data_
batch_normalization/renorm_mean/read
batch_normalization/renorm_mean/read/Output_0/Data_
batch_normalization/renorm_mean_weight
batch_normalization/renorm_mean_weight/Output_0/Data_
batch_normalization/renorm_mean_weight/read
batch_normalization/renorm_mean_weight/read/Output_0/Data_
batch_normalization/renorm_stddev
batch_normalization/renorm_stddev/Output_0/Data_
batch_normalization/renorm_stddev/read
batch_normalization/renorm_stddev/read/Output_0/Data_
batch_normalization/renorm_stddev_weight
batch_normalization/renorm_stddev_weight/Output_0/Data_
batch_normalization/renorm_stddev_weight/read
batch_normalization/renorm_stddev_weight/read/Output_0/Data_
batch_normalization/sub/add_
batch_normalization/sub/add_/Output_0/Data_
batch_normalization/sub/negate_
batch_normalization/sub/negate_/Output_0/Data_
batch_normalization/sub/x
batch_normalization/sub/x/Output_0/Data_
batch_normalization/sub_1/add_
batch_normalization/sub_1/add_/Output_0/Data_
batch_normalization/sub_1/negate_
batch_normalization/sub_1/negate_/Output_0/Data_
batch_normalization/sub_1/x
batch_normalization/sub_1/x/Output_0/Data_
batch_normalization/sub_2/add_
batch_normalization/sub_2/add_/Output_0/Data_
batch_normalization/sub_2/negate_
batch_normalization/sub_2/negate_/Output_0/Data_
batch_normalization/truediv/mul_
batch_normalization/truediv/mul_/Output_0/Data_
batch_normalization/truediv/reciprocal_
batch_normalization/truediv/reciprocal_/Output_0/Data_
batch_normalization/truediv_1/mul_
batch_normalization/truediv_1/mul_/Output_0/Data_
batch_normalization/truediv_1/reciprocal_
batch_normalization/truediv_1/reciprocal_/Output_0/Data_
batch_normalization_1/beta
batch_normalization_1/beta/read
batch_normalization_1/cond/FusedBatchNorm_1/Switch_1
batch_normalization_1/cond/FusedBatchNorm_1/Switch_2
batch_normalization_1/cond/FusedBatchNorm_1/Switch_3
batch_normalization_1/cond/FusedBatchNorm_1/Switch_4
batch_normalization_1/cond/pred_id
batch_normalization_1/gamma
batch_normalization_1/gamma/read
batch_normalization_1/moving_mean
batch_normalization_1/moving_mean/read
batch_normalization_1/moving_variance
batch_normalization_1/moving_variance/read
batch_normalization_2/beta
batch_normalization_2/beta/read
batch_normalization_2/cond/FusedBatchNorm_1/Switch_1
batch_normalization_2/cond/FusedBatchNorm_1/Switch_2
batch_normalization_2/cond/FusedBatchNorm_1/Switch_3
batch_normalization_2/cond/FusedBatchNorm_1/Switch_4
batch_normalization_2/cond/pred_id
batch_normalization_2/gamma
batch_normalization_2/gamma/read
batch_normalization_2/moving_mean
batch_normalization_2/moving_mean/read
batch_normalization_2/moving_variance
batch_normalization_2/moving_variance/read
conv2d/bias
conv2d/bias/read
conv2d/kernel
conv2d/kernel/read
conv_1/LeakyRelu/alpha
conv_1/bias
conv_1/bias/read
conv_1/kernel
conv_1/kernel/read
conv_2/LeakyRelu/alpha
conv_2/bias
conv_2/bias/read
conv_2/kernel
conv_2/kernel/read
pre_rnn_conv/LeakyRelu/alpha
pre_rnn_conv/bias
pre_rnn_conv/bias/read
pre_rnn_conv/kernel
pre_rnn_conv/kernel/read
reshape_to_rnn/shape
rnn/stack_bidirectional_rnn/cell_0/ReverseV2/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/ReverseV2/axis
rnn/stack_bidirectional_rnn/cell_0/ReverseV2/axis/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/ReverseV2/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/ReverseV2/axis
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/ReverseV2/axis/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/GRUCellZeroState/Const
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/GRUCellZeroState/ExpandDims
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/GRUCellZeroState/ExpandDims/dim
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/GRUCellZeroState/concat
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/GRUCellZeroState/concat/axis
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/GRUCellZeroState/concat/axis/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/GRUCellZeroState/zeros/Const
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/Maximum
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/Maximum/x
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/Minimum
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/Rank
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/Rank_1
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/Shape
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/Shape_1
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArrayUnstack/Shape
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArrayUnstack/Shape/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/concat
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/concat/axis
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/concat/axis/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/concat/values_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/concat_2
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/concat_2/axis
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/concat_2/axis/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/concat_2/values_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/range
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/range/delta
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/range/start
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/range_1
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/range_1/delta
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/range_1/start
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/strided_slice
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/strided_slice/stack
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/strided_slice/stack_1
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/strided_slice/stack_2
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/strided_slice_1
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/strided_slice_1/stack
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/strided_slice_1/stack_1
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/strided_slice_1/stack_2
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Enter_3
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Identity/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/NextIteration/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/NextIteration_1/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/NextIteration_2/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/BiasAdd/Enter
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/BiasAdd_1/Enter
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/MatMul/Enter
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/MatMul_1/Enter
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/concat/axis
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/concat/axis/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/concat_1/axis
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/concat_1/axis/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/split/split_dim
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/split/split_dim/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/sub/x
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/gru_cell/candidate/bias
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/gru_cell/candidate/bias/read
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/gru_cell/candidate/kernel
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/gru_cell/candidate/kernel/read
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/gru_cell/gates/bias
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/gru_cell/gates/bias/read
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/gru_cell/gates/kernel
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/gru_cell/gates/kernel/read
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/GRUCellZeroState/Const
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/GRUCellZeroState/ExpandDims
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/GRUCellZeroState/ExpandDims/dim
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/GRUCellZeroState/concat
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/GRUCellZeroState/concat/axis
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/GRUCellZeroState/concat/axis/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/GRUCellZeroState/zeros/Const
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/Maximum
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/Maximum/x
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/Minimum
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/Rank
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/Rank_1
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/Shape
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/Shape_1
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArrayUnstack/Shape
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArrayUnstack/Shape/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/concat
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/concat/axis
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/concat/axis/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/concat/values_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/concat_2
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/concat_2/axis
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/concat_2/axis/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/concat_2/values_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/range
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/range/delta
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/range/start
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/range_1
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/range_1/delta
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/range_1/start
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/strided_slice
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/strided_slice/stack
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/strided_slice/stack_1
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/strided_slice/stack_2
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/strided_slice_1
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/strided_slice_1/stack
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/strided_slice_1/stack_1
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/strided_slice_1/stack_2
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Enter_3
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Identity/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/NextIteration/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/NextIteration_1/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/NextIteration_2/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/BiasAdd/Enter
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/BiasAdd_1/Enter
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/MatMul/Enter
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/MatMul_1/Enter
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/concat/axis
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/concat/axis/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/concat_1/axis
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/concat_1/axis/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/split/split_dim
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/split/split_dim/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/sub/x
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/gru_cell/candidate/bias
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/gru_cell/candidate/bias/read
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/gru_cell/candidate/kernel
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/gru_cell/candidate/kernel/read
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/gru_cell/gates/bias
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/gru_cell/gates/bias/read
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/gru_cell/gates/kernel
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/gru_cell/gates/kernel/read
rnn/stack_bidirectional_rnn/cell_0/concat/axis
rnn/stack_bidirectional_rnn/cell_0/concat/axis/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/ReverseV2/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/ReverseV2/axis
rnn/stack_bidirectional_rnn/cell_1/ReverseV2/axis/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/ReverseV2/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/ReverseV2/axis
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/ReverseV2/axis/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/GRUCellZeroState/Const
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/GRUCellZeroState/ExpandDims
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/GRUCellZeroState/ExpandDims/dim
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/GRUCellZeroState/concat
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/GRUCellZeroState/concat/axis
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/GRUCellZeroState/concat/axis/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/GRUCellZeroState/zeros/Const
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/Maximum
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/Maximum/x
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/Minimum
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/Rank
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/Rank_1
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/Shape
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/Shape_1
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArrayUnstack/Shape
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArrayUnstack/Shape/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/concat
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/concat/axis
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/concat/axis/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/concat/values_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/concat_2
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/concat_2/axis
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/concat_2/axis/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/concat_2/values_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/range
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/range/delta
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/range/start
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/range_1
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/range_1/delta
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/range_1/start
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/strided_slice
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/strided_slice/stack
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/strided_slice/stack_1
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/strided_slice/stack_2
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/strided_slice_1
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/strided_slice_1/stack
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/strided_slice_1/stack_1
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/strided_slice_1/stack_2
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Enter_3
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Identity/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/NextIteration/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/NextIteration_1/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/NextIteration_2/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/BiasAdd/Enter
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/BiasAdd_1/Enter
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/MatMul/Enter
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/MatMul_1/Enter
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/concat/axis
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/concat/axis/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/concat_1/axis
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/concat_1/axis/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/split/split_dim
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/split/split_dim/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/sub/x
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/gru_cell/candidate/bias
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/gru_cell/candidate/bias/read
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/gru_cell/candidate/kernel
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/gru_cell/candidate/kernel/read
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/gru_cell/gates/bias
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/gru_cell/gates/bias/read
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/gru_cell/gates/kernel
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/gru_cell/gates/kernel/read
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/GRUCellZeroState/Const
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/GRUCellZeroState/ExpandDims
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/GRUCellZeroState/ExpandDims/dim
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/GRUCellZeroState/concat
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/GRUCellZeroState/concat/axis
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/GRUCellZeroState/concat/axis/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/GRUCellZeroState/zeros/Const
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/Maximum
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/Maximum/x
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/Minimum
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/Rank
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/Rank_1
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/Shape
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/Shape_1
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArrayUnstack/Shape
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArrayUnstack/Shape/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/concat
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/concat/axis
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/concat/axis/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/concat/values_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/concat_2
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/concat_2/axis
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/concat_2/axis/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/concat_2/values_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/range
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/range/delta
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/range/start
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/range_1
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/range_1/delta
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/range_1/start
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/strided_slice
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/strided_slice/stack
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/strided_slice/stack_1
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/strided_slice/stack_2
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/strided_slice_1
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/strided_slice_1/stack
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/strided_slice_1/stack_1
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/strided_slice_1/stack_2
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Enter_3
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Identity/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/NextIteration/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/NextIteration_1/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/NextIteration_2/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/BiasAdd/Enter
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/BiasAdd_1/Enter
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/MatMul/Enter
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/MatMul_1/Enter
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/concat/axis
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/concat/axis/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/concat_1/axis
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/concat_1/axis/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/split/split_dim
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/split/split_dim/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/sub/x
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/gru_cell/candidate/bias
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/gru_cell/candidate/bias/read
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/gru_cell/candidate/kernel
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/gru_cell/candidate/kernel/read
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/gru_cell/gates/bias
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/gru_cell/gates/bias/read
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/gru_cell/gates/kernel
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/gru_cell/gates/kernel/read
rnn/stack_bidirectional_rnn/cell_1/concat/axis
rnn/stack_bidirectional_rnn/cell_1/concat/axis/Output_0/Data_
split/split_dim
split/split_dim/Output_0/Data_
[ 2019-08-16 11:41:49,091 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.TensorIteratorInput.SimpleInputMatcher'>
[ 2019-08-16 11:41:49,110 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.TensorIteratorInput.BackEdgeSimpleInputMatcher'>
[ 2019-08-16 11:41:49,122 ] [ DEBUG ] [ TensorIteratorInput:233 ]  ================== SimpleBackEdgeInputFind ===============
[ 2019-08-16 11:41:49,122 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'op': 'TensorIteratorInput', 'axis': None, 'start': None, 'end': None, 'stride': None, 'part_size': None, 'in_ports_count': 3, 'out_ports_count': 1, 'infer': <function TensorIteratorInput.input_infer at 0x7fe50bcf7158>, 'external_port_id': None, 'internal_layer_id': None, 'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/Identity_3/TensorIteratorBackEdge_/TensorIteratorInput_', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a39ec80>), 'name', 'precision', 'type'], [('data', ['external_port_id', 'internal_layer_id', 'internal_port_id', 'axis', 'start', 'stride', 'part_size'], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}, 2: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,122 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'kind': 'data', 'precision': 'FP32', 'name': '1628', 'value': 'None', 'shape': None, 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], '_in_ports': {}, '_out_ports': {}}
[ 2019-08-16 11:41:49,122 ] [ DEBUG ] [ TensorIteratorInput:233 ]  ================== SimpleBackEdgeInputFind ===============
[ 2019-08-16 11:41:49,123 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'op': 'TensorIteratorInput', 'axis': None, 'start': None, 'end': None, 'stride': None, 'part_size': None, 'in_ports_count': 3, 'out_ports_count': 1, 'infer': <function TensorIteratorInput.input_infer at 0x7fe50bcf7158>, 'external_port_id': None, 'internal_layer_id': None, 'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/Identity_3/TensorIteratorBackEdge_/TensorIteratorInput_', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a3a3598>), 'name', 'precision', 'type'], [('data', ['external_port_id', 'internal_layer_id', 'internal_port_id', 'axis', 'start', 'stride', 'part_size'], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}, 2: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,123 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'kind': 'data', 'precision': 'FP32', 'name': '1630', 'value': 'None', 'shape': None, 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], '_in_ports': {}, '_out_ports': {}}
[ 2019-08-16 11:41:49,123 ] [ DEBUG ] [ TensorIteratorInput:233 ]  ================== SimpleBackEdgeInputFind ===============
[ 2019-08-16 11:41:49,123 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'op': 'TensorIteratorInput', 'axis': None, 'start': None, 'end': None, 'stride': None, 'part_size': None, 'in_ports_count': 3, 'out_ports_count': 1, 'infer': <function TensorIteratorInput.input_infer at 0x7fe50bcf7158>, 'external_port_id': None, 'internal_layer_id': None, 'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/Identity_3/TensorIteratorBackEdge_/TensorIteratorInput_', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a4b09d8>), 'name', 'precision', 'type'], [('data', ['external_port_id', 'internal_layer_id', 'internal_port_id', 'axis', 'start', 'stride', 'part_size'], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}, 2: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,123 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'kind': 'data', 'precision': 'FP32', 'name': '1632', 'value': 'None', 'shape': None, 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], '_in_ports': {}, '_out_ports': {}}
[ 2019-08-16 11:41:49,123 ] [ DEBUG ] [ TensorIteratorInput:233 ]  ================== SimpleBackEdgeInputFind ===============
[ 2019-08-16 11:41:49,124 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'op': 'TensorIteratorInput', 'axis': None, 'start': None, 'end': None, 'stride': None, 'part_size': None, 'in_ports_count': 3, 'out_ports_count': 1, 'infer': <function TensorIteratorInput.input_infer at 0x7fe50bcf7158>, 'external_port_id': None, 'internal_layer_id': None, 'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/Identity_3/TensorIteratorBackEdge_/TensorIteratorInput_', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a4d2e18>), 'name', 'precision', 'type'], [('data', ['external_port_id', 'internal_layer_id', 'internal_port_id', 'axis', 'start', 'stride', 'part_size'], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}, 2: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,124 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'kind': 'data', 'precision': 'FP32', 'name': '1634', 'value': 'None', 'shape': None, 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], '_in_ports': {}, '_out_ports': {}}
[ 2019-08-16 11:41:49,131 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.TensorIteratorMerge.TensorIteratorMerge'>
[ 2019-08-16 11:41:49,147 ] [ DEBUG ] [ graph:947 ]  Sink: rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0 for node rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_
[ 2019-08-16 11:41:49,147 ] [ DEBUG ] [ graph:948 ]  {'precision': 'FP32', 'kind': 'op', 'op': 'Result', 'type': 'Result', 'infer': <function Result.__init__.<locals>.<lambda> at 0x7fe50a3a32f0>, 'value': None, 'data_type': None, 'in_ports_count': 1, 'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a2c21e0>), 'name', 'precision', 'type'], [('data', [], []), '@ports', '@consts'])], '_in_ports': {0: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,147 ] [ DEBUG ] [ graph:949 ]  Add edge from rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,148 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'type': 'Const', 'op': 'Const', 'infer': <function tf_const_infer at 0x7fe511b10730>, 'value': '[-1 96]', 'shape': array([2]), 'data_type': None, 'out_ports_count': 1, 'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/TensorArrayReadV3/Output_0/Data_/ReshapeDim', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a4d2158>), 'name', 'precision', 'type'], [('data', ['offset', 'size'], []), '@ports', '@consts'])], '_in_ports': {}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,148 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'kind': 'data', 'precision': 'FP32', 'name': '51', 'value': '[-1 96]', 'shape': array([2]), 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], '_in_ports': {}, '_out_ports': {}}
[ 2019-08-16 11:41:49,148 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'type': 'Reshape', 'op': 'Reshape', 'reinterp_shape': True, 'in_ports_count': 2, 'out_ports_count': 1, 'infer': <function Reshape.infer at 0x7fe511aaa048>, 'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/TensorArrayReadV3/Output_0/Data_/InputSqueeze', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a37fbf8>), 'name', 'precision', 'type'], [('data', [], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,148 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/TensorArrayReadV3/Output_0/Data_', 'kind': 'data', 'shape': array([ 1, 96]), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/TensorArrayReadV3', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True, 'prev': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/TensorArrayReadV3/TensorIteratorInput_'}
[ 2019-08-16 11:41:49,149 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'type': 'Const', 'op': 'Const', 'infer': <function tf_const_infer at 0x7fe511b10730>, 'value': '[ 1 -1 96]', 'shape': array([3]), 'data_type': None, 'out_ports_count': 1, 'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/ReshapeDim', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a4ca730>), 'name', 'precision', 'type'], [('data', ['offset', 'size'], []), '@ports', '@consts'])], '_in_ports': {}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,149 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'kind': 'data', 'precision': 'FP32', 'name': '54', 'value': '[ 1 -1 96]', 'shape': array([3]), 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], '_in_ports': {}, '_out_ports': {}}
[ 2019-08-16 11:41:49,149 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'type': 'Reshape', 'op': 'Reshape', 'reinterp_shape': True, 'in_ports_count': 2, 'out_ports_count': 1, 'infer': <function Reshape.infer at 0x7fe511aaa048>, 'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/OutputUnsqueeze', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a444ea0>), 'name', 'precision', 'type'], [('data', [], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,149 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'kind': 'data', 'precision': 'FP32', 'name': '56', 'value': 'None', 'shape': array([ 1,  1, 96]), 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], '_in_ports': {}, '_out_ports': {}}
[ 2019-08-16 11:41:49,149 ] [ DEBUG ] [ graph:947 ]  Sink: 56/sink_port_0 for node 56
[ 2019-08-16 11:41:49,149 ] [ DEBUG ] [ graph:948 ]  {'precision': 'FP32', 'kind': 'op', 'op': 'Result', 'type': 'Result', 'infer': <function Result.__init__.<locals>.<lambda> at 0x7fe50a37f510>, 'value': None, 'data_type': None, 'in_ports_count': 1, 'name': '56/sink_port_0', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a3b3598>), 'name', 'precision', 'type'], [('data', [], []), '@ports', '@consts'])], '_in_ports': {0: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,150 ] [ DEBUG ] [ graph:949 ]  Add edge from 56 to 56/sink_port_0
[ 2019-08-16 11:41:49,150 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'type': 'TensorIterator', 'op': 'TensorIterator', 'input_port_map': [{'external_port_id': 8, 'internal_layer_id': 9, 'internal_port_id': 10, 'axis': 0, 'stride': None, 'part_size': None, 'start': None, 'end': None}, {'external_port_id': 11, 'internal_layer_id': 2, 'internal_port_id': 3, 'axis': None, 'stride': None, 'part_size': None, 'start': None, 'end': None}, {'external_port_id': 11, 'internal_layer_id': 4, 'internal_port_id': 5, 'axis': None, 'stride': None, 'part_size': None, 'start': None, 'end': None}, {'external_port_id': 11, 'internal_layer_id': 6, 'internal_port_id': 7, 'axis': None, 'stride': None, 'part_size': None, 'start': None, 'end': None}], 'output_port_map': [{'external_port_id': 14, 'internal_layer_id': 12, 'internal_port_id': 13, 'axis': 0, 'stride': None, 'part_size': None, 'start': None, 'end': None}], 'back_edges': [{'from_layer': 0, 'from_port': 1, 'to_layer': 2, 'to_port': 3}, {'from_layer': 0, 'from_port': 1, 'to_layer': 4, 'to_port': 5}, {'from_layer': 0, 'from_port': 1, 'to_layer': 6, 'to_port': 7}], 'body': <mo.graph.graph.Graph object at 0x7fe50a346e48>, 'sub_graphs': ['body'], 'infer': <function TensorIterator.infer at 0x7fe50bcd5bf8>, 'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/LoopCond/TensorIteratorCondition_/TensorIterator', 'in_ports_count': 2, 'out_ports_count': 1, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function TensorIterator.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a3b3b70>), 'name', 'precision', 'type'], [('data', [], []), '@ports', ('port_map', [], [('@list', <function TensorIterator.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a4cff28>, ('input', ['external_port_id', 'internal_layer_id', 'internal_port_id', 'axis', 'start', 'stride', 'end', 'part_size'], [])), ('@list', <function TensorIterator.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a4cfea0>, ('output', ['external_port_id', 'internal_layer_id', 'internal_port_id', 'axis', 'start', 'stride', 'end', 'part_size'], []))]), ('back_edges', [], [('@list', <function TensorIterator.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a4cfd08>, ('edge', [('from-layer', 'from_layer'), ('from-port', 'from_port'), ('to-layer', 'to_layer'), ('to-port', 'to_port')], []))]), ('body', [], [('@network', 'body')])])], '_in_ports': {0: {}, 1: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,150 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArrayStack/TensorArrayGatherV3/Output_0/Data_', 'kind': 'data', 'shape': array([28,  1, 96]), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/TensorArrayStack/TensorArrayGatherV3', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True, 'prev': None}
[ 2019-08-16 11:41:49,153 ] [ DEBUG ] [ graph:947 ]  Sink: rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0 for node rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_
[ 2019-08-16 11:41:49,153 ] [ DEBUG ] [ graph:948 ]  {'precision': 'FP32', 'kind': 'op', 'op': 'Result', 'type': 'Result', 'infer': <function Result.__init__.<locals>.<lambda> at 0x7fe50a3b37b8>, 'value': None, 'data_type': None, 'in_ports_count': 1, 'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a4ca840>), 'name', 'precision', 'type'], [('data', [], []), '@ports', '@consts'])], '_in_ports': {0: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,153 ] [ DEBUG ] [ graph:949 ]  Add edge from rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,154 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'type': 'Const', 'op': 'Const', 'infer': <function tf_const_infer at 0x7fe511b10730>, 'value': '[ -1 192]', 'shape': array([2]), 'data_type': None, 'out_ports_count': 1, 'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/TensorArrayReadV3/Output_0/Data_/ReshapeDim', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a4d2e18>), 'name', 'precision', 'type'], [('data', ['offset', 'size'], []), '@ports', '@consts'])], '_in_ports': {}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,154 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'kind': 'data', 'precision': 'FP32', 'name': '51', 'value': '[ -1 192]', 'shape': array([2]), 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], '_in_ports': {}, '_out_ports': {}}
[ 2019-08-16 11:41:49,154 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'type': 'Reshape', 'op': 'Reshape', 'reinterp_shape': True, 'in_ports_count': 2, 'out_ports_count': 1, 'infer': <function Reshape.infer at 0x7fe511aaa048>, 'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/TensorArrayReadV3/Output_0/Data_/InputSqueeze', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a4b02f0>), 'name', 'precision', 'type'], [('data', [], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,154 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/TensorArrayReadV3/Output_0/Data_', 'kind': 'data', 'shape': array([  1, 192]), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/TensorArrayReadV3', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True, 'prev': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/TensorArrayReadV3/TensorIteratorInput_'}
[ 2019-08-16 11:41:49,155 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'type': 'Const', 'op': 'Const', 'infer': <function tf_const_infer at 0x7fe511b10730>, 'value': '[ 1 -1 96]', 'shape': array([3]), 'data_type': None, 'out_ports_count': 1, 'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/ReshapeDim', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a2c2d90>), 'name', 'precision', 'type'], [('data', ['offset', 'size'], []), '@ports', '@consts'])], '_in_ports': {}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,155 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'kind': 'data', 'precision': 'FP32', 'name': '54', 'value': '[ 1 -1 96]', 'shape': array([3]), 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], '_in_ports': {}, '_out_ports': {}}
[ 2019-08-16 11:41:49,155 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'type': 'Reshape', 'op': 'Reshape', 'reinterp_shape': True, 'in_ports_count': 2, 'out_ports_count': 1, 'infer': <function Reshape.infer at 0x7fe511aaa048>, 'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/OutputUnsqueeze', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a56a1e0>), 'name', 'precision', 'type'], [('data', [], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,155 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'kind': 'data', 'precision': 'FP32', 'name': '56', 'value': 'None', 'shape': array([ 1,  1, 96]), 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], '_in_ports': {}, '_out_ports': {}}
[ 2019-08-16 11:41:49,155 ] [ DEBUG ] [ graph:947 ]  Sink: 56/sink_port_0 for node 56
[ 2019-08-16 11:41:49,155 ] [ DEBUG ] [ graph:948 ]  {'precision': 'FP32', 'kind': 'op', 'op': 'Result', 'type': 'Result', 'infer': <function Result.__init__.<locals>.<lambda> at 0x7fe50a4480d0>, 'value': None, 'data_type': None, 'in_ports_count': 1, 'name': '56/sink_port_0', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a3a3510>), 'name', 'precision', 'type'], [('data', [], []), '@ports', '@consts'])], '_in_ports': {0: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,155 ] [ DEBUG ] [ graph:949 ]  Add edge from 56 to 56/sink_port_0
[ 2019-08-16 11:41:49,156 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'type': 'TensorIterator', 'op': 'TensorIterator', 'input_port_map': [{'external_port_id': 8, 'internal_layer_id': 9, 'internal_port_id': 10, 'axis': 0, 'stride': None, 'part_size': None, 'start': None, 'end': None}, {'external_port_id': 11, 'internal_layer_id': 2, 'internal_port_id': 3, 'axis': None, 'stride': None, 'part_size': None, 'start': None, 'end': None}, {'external_port_id': 11, 'internal_layer_id': 4, 'internal_port_id': 5, 'axis': None, 'stride': None, 'part_size': None, 'start': None, 'end': None}, {'external_port_id': 11, 'internal_layer_id': 6, 'internal_port_id': 7, 'axis': None, 'stride': None, 'part_size': None, 'start': None, 'end': None}], 'output_port_map': [{'external_port_id': 14, 'internal_layer_id': 12, 'internal_port_id': 13, 'axis': 0, 'stride': None, 'part_size': None, 'start': None, 'end': None}], 'back_edges': [{'from_layer': 0, 'from_port': 1, 'to_layer': 2, 'to_port': 3}, {'from_layer': 0, 'from_port': 1, 'to_layer': 4, 'to_port': 5}, {'from_layer': 0, 'from_port': 1, 'to_layer': 6, 'to_port': 7}], 'body': <mo.graph.graph.Graph object at 0x7fe50a4047b8>, 'sub_graphs': ['body'], 'infer': <function TensorIterator.infer at 0x7fe50bcd5bf8>, 'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/LoopCond/TensorIteratorCondition_/TensorIterator', 'in_ports_count': 2, 'out_ports_count': 1, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function TensorIterator.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a4e8268>), 'name', 'precision', 'type'], [('data', [], []), '@ports', ('port_map', [], [('@list', <function TensorIterator.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a4e82f0>, ('input', ['external_port_id', 'internal_layer_id', 'internal_port_id', 'axis', 'start', 'stride', 'end', 'part_size'], [])), ('@list', <function TensorIterator.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a4e81e0>, ('output', ['external_port_id', 'internal_layer_id', 'internal_port_id', 'axis', 'start', 'stride', 'end', 'part_size'], []))]), ('back_edges', [], [('@list', <function TensorIterator.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a4e8158>, ('edge', [('from-layer', 'from_layer'), ('from-port', 'from_port'), ('to-layer', 'to_layer'), ('to-port', 'to_port')], []))]), ('body', [], [('@network', 'body')])])], '_in_ports': {0: {}, 1: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,156 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArrayStack/TensorArrayGatherV3/Output_0/Data_', 'kind': 'data', 'shape': array([28,  1, 96]), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/TensorArrayStack/TensorArrayGatherV3', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True, 'prev': None}
[ 2019-08-16 11:41:49,159 ] [ DEBUG ] [ graph:947 ]  Sink: rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0 for node rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_
[ 2019-08-16 11:41:49,159 ] [ DEBUG ] [ graph:948 ]  {'precision': 'FP32', 'kind': 'op', 'op': 'Result', 'type': 'Result', 'infer': <function Result.__init__.<locals>.<lambda> at 0x7fe50a4b0510>, 'value': None, 'data_type': None, 'in_ports_count': 1, 'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a4caae8>), 'name', 'precision', 'type'], [('data', [], []), '@ports', '@consts'])], '_in_ports': {0: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,159 ] [ DEBUG ] [ graph:949 ]  Add edge from rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,159 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'type': 'Const', 'op': 'Const', 'infer': <function tf_const_infer at 0x7fe511b10730>, 'value': '[ -1 192]', 'shape': array([2]), 'data_type': None, 'out_ports_count': 1, 'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/TensorArrayReadV3/Output_0/Data_/ReshapeDim', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a3b36a8>), 'name', 'precision', 'type'], [('data', ['offset', 'size'], []), '@ports', '@consts'])], '_in_ports': {}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,160 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'kind': 'data', 'precision': 'FP32', 'name': '51', 'value': '[ -1 192]', 'shape': array([2]), 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], '_in_ports': {}, '_out_ports': {}}
[ 2019-08-16 11:41:49,160 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'type': 'Reshape', 'op': 'Reshape', 'reinterp_shape': True, 'in_ports_count': 2, 'out_ports_count': 1, 'infer': <function Reshape.infer at 0x7fe511aaa048>, 'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/TensorArrayReadV3/Output_0/Data_/InputSqueeze', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a3b3950>), 'name', 'precision', 'type'], [('data', [], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,160 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/TensorArrayReadV3/Output_0/Data_', 'kind': 'data', 'shape': array([  1, 192]), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/TensorArrayReadV3', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True, 'prev': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/TensorArrayReadV3/TensorIteratorInput_'}
[ 2019-08-16 11:41:49,160 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'type': 'Const', 'op': 'Const', 'infer': <function tf_const_infer at 0x7fe511b10730>, 'value': '[ 1 -1 96]', 'shape': array([3]), 'data_type': None, 'out_ports_count': 1, 'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/ReshapeDim', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a4b5b70>), 'name', 'precision', 'type'], [('data', ['offset', 'size'], []), '@ports', '@consts'])], '_in_ports': {}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,161 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'kind': 'data', 'precision': 'FP32', 'name': '54', 'value': '[ 1 -1 96]', 'shape': array([3]), 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], '_in_ports': {}, '_out_ports': {}}
[ 2019-08-16 11:41:49,161 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'type': 'Reshape', 'op': 'Reshape', 'reinterp_shape': True, 'in_ports_count': 2, 'out_ports_count': 1, 'infer': <function Reshape.infer at 0x7fe511aaa048>, 'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/OutputUnsqueeze', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a4b5d08>), 'name', 'precision', 'type'], [('data', [], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,161 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'kind': 'data', 'precision': 'FP32', 'name': '56', 'value': 'None', 'shape': array([ 1,  1, 96]), 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], '_in_ports': {}, '_out_ports': {}}
[ 2019-08-16 11:41:49,161 ] [ DEBUG ] [ graph:947 ]  Sink: 56/sink_port_0 for node 56
[ 2019-08-16 11:41:49,161 ] [ DEBUG ] [ graph:948 ]  {'precision': 'FP32', 'kind': 'op', 'op': 'Result', 'type': 'Result', 'infer': <function Result.__init__.<locals>.<lambda> at 0x7fe50a4b5ea0>, 'value': None, 'data_type': None, 'in_ports_count': 1, 'name': '56/sink_port_0', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a3a3400>), 'name', 'precision', 'type'], [('data', [], []), '@ports', '@consts'])], '_in_ports': {0: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,161 ] [ DEBUG ] [ graph:949 ]  Add edge from 56 to 56/sink_port_0
[ 2019-08-16 11:41:49,162 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'type': 'TensorIterator', 'op': 'TensorIterator', 'input_port_map': [{'external_port_id': 8, 'internal_layer_id': 9, 'internal_port_id': 10, 'axis': 0, 'stride': None, 'part_size': None, 'start': None, 'end': None}, {'external_port_id': 11, 'internal_layer_id': 2, 'internal_port_id': 3, 'axis': None, 'stride': None, 'part_size': None, 'start': None, 'end': None}, {'external_port_id': 11, 'internal_layer_id': 4, 'internal_port_id': 5, 'axis': None, 'stride': None, 'part_size': None, 'start': None, 'end': None}, {'external_port_id': 11, 'internal_layer_id': 6, 'internal_port_id': 7, 'axis': None, 'stride': None, 'part_size': None, 'start': None, 'end': None}], 'output_port_map': [{'external_port_id': 14, 'internal_layer_id': 12, 'internal_port_id': 13, 'axis': 0, 'stride': None, 'part_size': None, 'start': None, 'end': None}], 'back_edges': [{'from_layer': 0, 'from_port': 1, 'to_layer': 2, 'to_port': 3}, {'from_layer': 0, 'from_port': 1, 'to_layer': 4, 'to_port': 5}, {'from_layer': 0, 'from_port': 1, 'to_layer': 6, 'to_port': 7}], 'body': <mo.graph.graph.Graph object at 0x7fe50a4046a0>, 'sub_graphs': ['body'], 'infer': <function TensorIterator.infer at 0x7fe50bcd5bf8>, 'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/LoopCond/TensorIteratorCondition_/TensorIterator', 'in_ports_count': 2, 'out_ports_count': 1, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function TensorIterator.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a4b58c8>), 'name', 'precision', 'type'], [('data', [], []), '@ports', ('port_map', [], [('@list', <function TensorIterator.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a4cfe18>, ('input', ['external_port_id', 'internal_layer_id', 'internal_port_id', 'axis', 'start', 'stride', 'end', 'part_size'], [])), ('@list', <function TensorIterator.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a4cf2f0>, ('output', ['external_port_id', 'internal_layer_id', 'internal_port_id', 'axis', 'start', 'stride', 'end', 'part_size'], []))]), ('back_edges', [], [('@list', <function TensorIterator.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a4cf048>, ('edge', [('from-layer', 'from_layer'), ('from-port', 'from_port'), ('to-layer', 'to_layer'), ('to-port', 'to_port')], []))]), ('body', [], [('@network', 'body')])])], '_in_ports': {0: {}, 1: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,162 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArrayStack/TensorArrayGatherV3/Output_0/Data_', 'kind': 'data', 'shape': array([28,  1, 96]), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/TensorArrayStack/TensorArrayGatherV3', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True, 'prev': None}
[ 2019-08-16 11:41:49,165 ] [ DEBUG ] [ graph:947 ]  Sink: rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0 for node rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_
[ 2019-08-16 11:41:49,165 ] [ DEBUG ] [ graph:948 ]  {'precision': 'FP32', 'kind': 'op', 'op': 'Result', 'type': 'Result', 'infer': <function Result.__init__.<locals>.<lambda> at 0x7fe50a4b09d8>, 'value': None, 'data_type': None, 'in_ports_count': 1, 'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a4ca1e0>), 'name', 'precision', 'type'], [('data', [], []), '@ports', '@consts'])], '_in_ports': {0: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,165 ] [ DEBUG ] [ graph:949 ]  Add edge from rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_ to rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,165 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'type': 'Const', 'op': 'Const', 'infer': <function tf_const_infer at 0x7fe511b10730>, 'value': '[-1 96]', 'shape': array([2]), 'data_type': None, 'out_ports_count': 1, 'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/TensorArrayReadV3/Output_0/Data_/ReshapeDim', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a3b31e0>), 'name', 'precision', 'type'], [('data', ['offset', 'size'], []), '@ports', '@consts'])], '_in_ports': {}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,166 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'kind': 'data', 'precision': 'FP32', 'name': '51', 'value': '[-1 96]', 'shape': array([2]), 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], '_in_ports': {}, '_out_ports': {}}
[ 2019-08-16 11:41:49,166 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'type': 'Reshape', 'op': 'Reshape', 'reinterp_shape': True, 'in_ports_count': 2, 'out_ports_count': 1, 'infer': <function Reshape.infer at 0x7fe511aaa048>, 'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/TensorArrayReadV3/Output_0/Data_/InputSqueeze', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a3b3378>), 'name', 'precision', 'type'], [('data', [], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,166 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/TensorArrayReadV3/Output_0/Data_', 'kind': 'data', 'shape': array([ 1, 96]), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/TensorArrayReadV3', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True, 'prev': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/TensorArrayReadV3/TensorIteratorInput_'}
[ 2019-08-16 11:41:49,166 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'type': 'Const', 'op': 'Const', 'infer': <function tf_const_infer at 0x7fe511b10730>, 'value': '[ 1 -1 96]', 'shape': array([3]), 'data_type': None, 'out_ports_count': 1, 'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/ReshapeDim', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a4b5950>), 'name', 'precision', 'type'], [('data', ['offset', 'size'], []), '@ports', '@consts'])], '_in_ports': {}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,167 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'kind': 'data', 'precision': 'FP32', 'name': '54', 'value': '[ 1 -1 96]', 'shape': array([3]), 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], '_in_ports': {}, '_out_ports': {}}
[ 2019-08-16 11:41:49,167 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'type': 'Reshape', 'op': 'Reshape', 'reinterp_shape': True, 'in_ports_count': 2, 'out_ports_count': 1, 'infer': <function Reshape.infer at 0x7fe511aaa048>, 'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/OutputUnsqueeze', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a4b5ae8>), 'name', 'precision', 'type'], [('data', [], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,167 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'kind': 'data', 'precision': 'FP32', 'name': '56', 'value': 'None', 'shape': array([ 1,  1, 96]), 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], '_in_ports': {}, '_out_ports': {}}
[ 2019-08-16 11:41:49,167 ] [ DEBUG ] [ graph:947 ]  Sink: 56/sink_port_0 for node 56
[ 2019-08-16 11:41:49,167 ] [ DEBUG ] [ graph:948 ]  {'precision': 'FP32', 'kind': 'op', 'op': 'Result', 'type': 'Result', 'infer': <function Result.__init__.<locals>.<lambda> at 0x7fe50a4b5400>, 'value': None, 'data_type': None, 'in_ports_count': 1, 'name': '56/sink_port_0', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a3a3268>), 'name', 'precision', 'type'], [('data', [], []), '@ports', '@consts'])], '_in_ports': {0: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,167 ] [ DEBUG ] [ graph:949 ]  Add edge from 56 to 56/sink_port_0
[ 2019-08-16 11:41:49,168 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'type': 'TensorIterator', 'op': 'TensorIterator', 'input_port_map': [{'external_port_id': 8, 'internal_layer_id': 9, 'internal_port_id': 10, 'axis': 0, 'stride': None, 'part_size': None, 'start': None, 'end': None}, {'external_port_id': 11, 'internal_layer_id': 2, 'internal_port_id': 3, 'axis': None, 'stride': None, 'part_size': None, 'start': None, 'end': None}, {'external_port_id': 11, 'internal_layer_id': 4, 'internal_port_id': 5, 'axis': None, 'stride': None, 'part_size': None, 'start': None, 'end': None}, {'external_port_id': 11, 'internal_layer_id': 6, 'internal_port_id': 7, 'axis': None, 'stride': None, 'part_size': None, 'start': None, 'end': None}], 'output_port_map': [{'external_port_id': 14, 'internal_layer_id': 12, 'internal_port_id': 13, 'axis': 0, 'stride': None, 'part_size': None, 'start': None, 'end': None}], 'back_edges': [{'from_layer': 0, 'from_port': 1, 'to_layer': 2, 'to_port': 3}, {'from_layer': 0, 'from_port': 1, 'to_layer': 4, 'to_port': 5}, {'from_layer': 0, 'from_port': 1, 'to_layer': 6, 'to_port': 7}], 'body': <mo.graph.graph.Graph object at 0x7fe50a346908>, 'sub_graphs': ['body'], 'infer': <function TensorIterator.infer at 0x7fe50bcd5bf8>, 'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/LoopCond/TensorIteratorCondition_/TensorIterator', 'in_ports_count': 2, 'out_ports_count': 1, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function TensorIterator.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a4d20d0>), 'name', 'precision', 'type'], [('data', [], []), '@ports', ('port_map', [], [('@list', <function TensorIterator.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a4d22f0>, ('input', ['external_port_id', 'internal_layer_id', 'internal_port_id', 'axis', 'start', 'stride', 'end', 'part_size'], [])), ('@list', <function TensorIterator.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a4d2840>, ('output', ['external_port_id', 'internal_layer_id', 'internal_port_id', 'axis', 'start', 'stride', 'end', 'part_size'], []))]), ('back_edges', [], [('@list', <function TensorIterator.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a3d26a8>, ('edge', [('from-layer', 'from_layer'), ('from-port', 'from_port'), ('to-layer', 'to_layer'), ('to-port', 'to_port')], []))]), ('body', [], [('@network', 'body')])])], '_in_ports': {0: {}, 1: {}}, '_out_ports': {0: {}}}
[ 2019-08-16 11:41:49,168 ] [ DEBUG ] [ op:223 ]  Finished running infer function, data nodes attributes: {'name': 'rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArrayStack/TensorArrayGatherV3/Output_0/Data_', 'kind': 'data', 'shape': array([28,  1, 96]), 'value': 'None', 'data_type': None, 'infer': None, 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'fw_tensor_debug_info': [('rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/TensorArrayStack/TensorArrayGatherV3', 0)], 'is_output_reachable': True, 'is_undead': False, 'is_const_producer': False, 'is_partial_inferred': True, 'executable': True, 'prev': None}
[ 2019-08-16 11:41:49,188 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.CheckForCycle.CheckForCycle'>
[ 2019-08-16 11:41:49,193 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
seq_len/sink_port_0
[ 2019-08-16 11:41:49,209 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1580
1581
1582
1583
1584
1590
1591
1592
1593
1594
1600
1601
1602
1603
1604
1610
1611
1612
1613
1614
1615
1616
1617
1620
1623
1624
1625
1626
761/Output_0/Data_
768/Output_0/Data_
769/Output_0/Data_
770/Output_0/Data_
771/Output_0/Data_
772/Output_0/Data_
PlaceholderWithDefault/Output_0/Data_
PlaceholderWithDefault/input/Output_0/Data_
batch_normalization/batchnorm/add/Output_0/Data_
batch_normalization/batchnorm/add/y/Output_0/Data_
batch_normalization/beta/Output_0/Data_
batch_normalization/cond/Switch_2/Output_0/Data_
batch_normalization/cond/pred_id/Output_0/Data_
batch_normalization/cond_1/Merge/Output_0/Data_
batch_normalization/cond_1/Switch_2/Output_0/Data_
batch_normalization/cond_1/pred_id/Output_0/Data_
batch_normalization/cond_2/ones_like/Const/Output_0/Data_
batch_normalization/cond_2/ones_like/Shape/Output_0/Data_
batch_normalization/cond_3/zeros_like/Output_0/Data_
batch_normalization/gamma/Output_0/Data_
batch_normalization/moving_mean/Output_0/Data_
batch_normalization/moving_mean/read/Output_0/Data_
batch_normalization/moving_variance/Output_0/Data_
batch_normalization/moving_variance/read/Output_0/Data_
batch_normalization_1/beta/Output_0/Data_
batch_normalization_1/beta/read/Output_0/Data_
batch_normalization_1/gamma/Output_0/Data_
batch_normalization_1/gamma/read/Output_0/Data_
batch_normalization_1/moving_mean/Output_0/Data_
batch_normalization_1/moving_mean/read/Output_0/Data_
batch_normalization_1/moving_variance/Output_0/Data_
batch_normalization_1/moving_variance/read/Output_0/Data_
batch_normalization_2/beta/Output_0/Data_
batch_normalization_2/beta/read/Output_0/Data_
batch_normalization_2/gamma/Output_0/Data_
batch_normalization_2/gamma/read/Output_0/Data_
batch_normalization_2/moving_mean/Output_0/Data_
batch_normalization_2/moving_mean/read/Output_0/Data_
batch_normalization_2/moving_variance/Output_0/Data_
batch_normalization_2/moving_variance/read/Output_0/Data_
conv2d/bias/Output_0/Data_
conv2d/kernel/Output_0/Data_
conv_1/bias/Output_0/Data_
conv_1/kernel/Output_0/Data_
conv_2/bias/Output_0/Data_
conv_2/kernel/Output_0/Data_
pre_rnn_conv/bias/Output_0/Data_
pre_rnn_conv/kernel/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/GRUCellZeroState/Const/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/GRUCellZeroState/ExpandDims/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/GRUCellZeroState/ExpandDims/dim/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/GRUCellZeroState/concat/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/GRUCellZeroState/zeros/Const/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/Maximum/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/Maximum/x/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/Minimum/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/Rank/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/Rank_1/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/Shape/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/Shape_1/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/concat/values_0/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/concat_2/values_0/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/range/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/range/delta/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/range/start/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/range_1/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/range_1/delta/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/range_1/start/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/strided_slice/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/strided_slice/stack/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/strided_slice/stack_1/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/strided_slice/stack_2/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/strided_slice_1/stack/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/strided_slice_1/stack_1/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/strided_slice_1/stack_2/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/gru_cell/candidate/bias/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/gru_cell/candidate/bias/read/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/gru_cell/candidate/kernel/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/gru_cell/candidate/kernel/read/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/gru_cell/gates/bias/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/gru_cell/gates/bias/read/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/gru_cell/gates/kernel/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/gru_cell/gates/kernel/read/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/GRUCellZeroState/Const/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/GRUCellZeroState/ExpandDims/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/GRUCellZeroState/ExpandDims/dim/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/GRUCellZeroState/concat/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/GRUCellZeroState/zeros/Const/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/Maximum/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/Maximum/x/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/Minimum/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/Rank/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/Rank_1/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/Shape/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/Shape_1/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/concat/values_0/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/concat_2/values_0/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/range/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/range/delta/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/range/start/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/range_1/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/range_1/delta/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/range_1/start/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/strided_slice/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/strided_slice/stack/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/strided_slice/stack_1/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/strided_slice/stack_2/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/strided_slice_1/stack/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/strided_slice_1/stack_1/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/strided_slice_1/stack_2/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/gru_cell/candidate/bias/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/gru_cell/candidate/bias/read/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/gru_cell/candidate/kernel/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/gru_cell/candidate/kernel/read/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/gru_cell/gates/bias/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/gru_cell/gates/bias/read/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/gru_cell/gates/kernel/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/gru_cell/gates/kernel/read/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/GRUCellZeroState/Const/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/GRUCellZeroState/ExpandDims/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/GRUCellZeroState/ExpandDims/dim/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/GRUCellZeroState/concat/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/GRUCellZeroState/zeros/Const/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/Maximum/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/Maximum/x/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/Minimum/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/Rank/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/Rank_1/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/Shape/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/Shape_1/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/concat/values_0/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/concat_2/values_0/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/range/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/range/delta/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/range/start/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/range_1/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/range_1/delta/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/range_1/start/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/strided_slice/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/strided_slice/stack/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/strided_slice/stack_1/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/strided_slice/stack_2/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/strided_slice_1/stack/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/strided_slice_1/stack_1/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/strided_slice_1/stack_2/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/gru_cell/candidate/bias/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/gru_cell/candidate/bias/read/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/gru_cell/candidate/kernel/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/gru_cell/candidate/kernel/read/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/gru_cell/gates/bias/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/gru_cell/gates/bias/read/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/gru_cell/gates/kernel/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/gru_cell/gates/kernel/read/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/GRUCellZeroState/Const/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/GRUCellZeroState/ExpandDims/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/GRUCellZeroState/ExpandDims/dim/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/GRUCellZeroState/concat/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/GRUCellZeroState/zeros/Const/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/Maximum/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/Maximum/x/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/Minimum/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/Rank/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/Rank_1/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/Shape/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/Shape_1/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/concat/values_0/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/concat_2/values_0/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/range/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/range/delta/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/range/start/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/range_1/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/range_1/delta/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/range_1/start/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/strided_slice/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/strided_slice/stack/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/strided_slice/stack_1/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/strided_slice/stack_2/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/strided_slice_1/stack/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/strided_slice_1/stack_1/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/strided_slice_1/stack_2/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/gru_cell/candidate/bias/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/gru_cell/candidate/bias/read/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/gru_cell/candidate/kernel/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/gru_cell/candidate/kernel/read/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/gru_cell/gates/bias/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/gru_cell/gates/bias/read/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/gru_cell/gates/kernel/Output_0/Data_
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/gru_cell/gates/kernel/read/Output_0/Data_
[ 2019-08-16 11:41:49,217 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,219 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 1585
1586
1587
1588
1589
1619
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/TensorArrayReadV3/Output_0/Data_/ReshapeDim
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/ReshapeDim
[ 2019-08-16 11:41:49,221 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,223 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 1595
1596
1597
1598
1599
1621
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/TensorArrayReadV3/Output_0/Data_/ReshapeDim
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/ReshapeDim
[ 2019-08-16 11:41:49,225 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,228 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 1605
1606
1607
1608
1609
1618
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/TensorArrayReadV3/Output_0/Data_/ReshapeDim
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/ReshapeDim
[ 2019-08-16 11:41:49,229 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,232 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 1575
1576
1577
1578
1579
1622
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/TensorArrayReadV3/Output_0/Data_/ReshapeDim
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/ReshapeDim
[ 2019-08-16 11:41:49,240 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.SharedWeightsDuplication.SharedWeightsDuplication'>
[ 2019-08-16 11:41:49,249 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
seq_len/sink_port_0
[ 2019-08-16 11:41:49,260 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
batch_normalization/gamma/read/Output_0/Data_
[ 2019-08-16 11:41:49,267 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,270 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 58
59
60
61
62
63
64
65
[ 2019-08-16 11:41:49,271 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,274 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 58
59
60
61
62
63
64
65
[ 2019-08-16 11:41:49,276 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,279 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 58
59
60
61
62
63
64
65
[ 2019-08-16 11:41:49,280 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,283 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 58
59
60
61
62
63
64
65
[ 2019-08-16 11:41:49,291 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.pass_separator.PreMiddleStart'>
[ 2019-08-16 11:41:49,299 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.UselessSplitEraser.UselessSplitEraser'>
[ 2019-08-16 11:41:49,316 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.InputCut.MiddleInputCut'>
[ 2019-08-16 11:41:49,317 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
seq_len/sink_port_0
[ 2019-08-16 11:41:49,328 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
split
split/Output_0/Data_
[ 2019-08-16 11:41:49,335 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,338 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 66
67
68
69
70
71
72
73
[ 2019-08-16 11:41:49,339 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,342 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 66
67
68
69
70
71
72
73
[ 2019-08-16 11:41:49,343 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,346 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 66
67
68
69
70
71
72
73
[ 2019-08-16 11:41:49,347 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,350 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 66
67
68
69
70
71
72
73
[ 2019-08-16 11:41:49,358 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.ChangePlaceholderTypes.ChangePlaceholderTypes'>
[ 2019-08-16 11:41:49,360 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
seq_len/sink_port_0
[ 2019-08-16 11:41:49,371 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
[ 2019-08-16 11:41:49,378 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,381 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 74
75
76
77
78
79
80
81
[ 2019-08-16 11:41:49,382 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,385 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 74
75
76
77
78
79
80
81
[ 2019-08-16 11:41:49,386 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,389 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 74
75
76
77
78
79
80
81
[ 2019-08-16 11:41:49,390 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,393 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 74
75
76
77
78
79
80
81
[ 2019-08-16 11:41:49,402 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.ScaleInput.ScaleInput'>
[ 2019-08-16 11:41:49,419 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.AddMeanScaleValues.AddMeanScaleValues'>
[ 2019-08-16 11:41:49,427 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.RemoveIdentity.RemoveNodesWithZeroPhase'>
[ 2019-08-16 11:41:49,437 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
seq_len/sink_port_0
[ 2019-08-16 11:41:49,448 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
[ 2019-08-16 11:41:49,455 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,458 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 82
83
84
85
86
87
88
89
[ 2019-08-16 11:41:49,459 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,462 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 82
83
84
85
86
87
88
89
[ 2019-08-16 11:41:49,463 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,466 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 82
83
84
85
86
87
88
89
[ 2019-08-16 11:41:49,467 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,470 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 82
83
84
85
86
87
88
89
[ 2019-08-16 11:41:49,479 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.RemoveIdentity.RemoveDropout'>
[ 2019-08-16 11:41:49,496 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.L2NormToNorm.L2NormToNorm'>
[ 2019-08-16 11:41:49,509 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
seq_len/sink_port_0
[ 2019-08-16 11:41:49,520 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
[ 2019-08-16 11:41:49,527 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,530 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 90
91
92
93
94
95
96
97
[ 2019-08-16 11:41:49,531 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,535 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 90
91
92
93
94
95
96
97
[ 2019-08-16 11:41:49,536 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,539 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 90
91
92
93
94
95
96
97
[ 2019-08-16 11:41:49,540 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,543 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 90
91
92
93
94
95
96
97
[ 2019-08-16 11:41:49,552 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.DilatedConvolution.DilatedConvolutionConverter'>
[ 2019-08-16 11:41:49,562 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
seq_len/sink_port_0
[ 2019-08-16 11:41:49,573 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
[ 2019-08-16 11:41:49,580 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,583 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 100
101
102
103
104
105
98
99
[ 2019-08-16 11:41:49,584 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,587 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 100
101
102
103
104
105
98
99
[ 2019-08-16 11:41:49,589 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,592 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 100
101
102
103
104
105
98
99
[ 2019-08-16 11:41:49,593 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,596 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 100
101
102
103
104
105
98
99
[ 2019-08-16 11:41:49,604 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.CustomSubgraphCall.CustomSubgraphCall'>
[ 2019-08-16 11:41:49,612 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
seq_len/sink_port_0
[ 2019-08-16 11:41:49,624 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
[ 2019-08-16 11:41:49,631 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,634 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 106
107
108
109
110
111
112
113
[ 2019-08-16 11:41:49,635 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,638 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 106
107
108
109
110
111
112
113
[ 2019-08-16 11:41:49,639 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,642 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 106
107
108
109
110
111
112
113
[ 2019-08-16 11:41:49,643 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,646 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 106
107
108
109
110
111
112
113
[ 2019-08-16 11:41:49,655 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.ConvertMultiInputConv.ConvertMultiInputConv'>
[ 2019-08-16 11:41:49,664 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
seq_len/sink_port_0
[ 2019-08-16 11:41:49,676 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
[ 2019-08-16 11:41:49,683 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,686 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 114
115
116
117
118
119
120
121
[ 2019-08-16 11:41:49,687 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,690 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 114
115
116
117
118
119
120
121
[ 2019-08-16 11:41:49,691 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,694 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 114
115
116
117
118
119
120
121
[ 2019-08-16 11:41:49,695 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,698 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 114
115
116
117
118
119
120
121
[ 2019-08-16 11:41:49,707 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.Cast.CastToFloatMark'>
[ WARNING ]  Deleting Cast node seq_len to <class 'numpy.int32'> from network since Cast operation isn't supported yet. Inference results can be incorrect
[ 2019-08-16 11:41:49,724 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.RemoveIdentity.RemoveIdentity'>
[ 2019-08-16 11:41:49,729 ] [ DEBUG ] [ eliminate:239 ]  Removing op node: seq_len
[ 2019-08-16 11:41:49,729 ] [ DEBUG ] [ eliminate:239 ]  Removing op node: batch_normalization/renorm_r
[ 2019-08-16 11:41:49,729 ] [ DEBUG ] [ eliminate:239 ]  Removing op node: batch_normalization/renorm_d
[ 2019-08-16 11:41:49,729 ] [ DEBUG ] [ eliminate:239 ]  Removing op node: concat
[ 2019-08-16 11:41:49,741 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.pass_separator.MiddleStart'>
[ INFO ]  Skip replacer <class 'extensions.middle.UpsampleToResample.UpsampleToResample'> (graph_condition not satisfied)
[ 2019-08-16 11:41:49,749 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.TileReplacer.TileMultipleAxisReplacer'>
[ 2019-08-16 11:41:49,759 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
seq_len/sink_port_0
[ 2019-08-16 11:41:49,769 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
[ 2019-08-16 11:41:49,776 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,779 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 122
123
124
125
126
127
128
129
[ 2019-08-16 11:41:49,781 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,783 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 122
123
124
125
126
127
128
129
[ 2019-08-16 11:41:49,785 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,788 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 122
123
124
125
126
127
128
129
[ 2019-08-16 11:41:49,789 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:49,792 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 122
123
124
125
126
127
128
129
[ 2019-08-16 11:41:49,800 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.TF_lstm_cell_to_generic.TensorFlowLSTMtoGeneric'>
[ 2019-08-16 11:41:49,816 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.SwapAxesMiddleReplacer.SwapAxisMiddleReplacer'>
[ 2019-08-16 11:41:49,833 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.space_to_depth.SpaceToDepth'>
[ 2019-08-16 11:41:49,859 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.ShuffleChannel.ShuffleChannel'>
[ 2019-08-16 11:41:49,875 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.ReverseTransposeNormalization.ReverseTransposeNormalization'>
[ INFO ]  Skip replacer <class 'extensions.middle.ReplaceSpliceNodePattern.ReplaceSpliceNodePattern'> (enabled = False)
[ INFO ]  Skip replacer <class 'extensions.middle.ReplaceMemoryOffsetWithSplice.ReplaceMemoryOffsetWithMemoryNodePattern'> (enabled = False)
[ INFO ]  Skip replacer <class 'extensions.middle.ReplaceMemoryOffsetWithSplice.ReplaceMemoryOffsetNodePattern'> (enabled = False)
[ INFO ]  Skip replacer <class 'extensions.middle.RemoveUselessCrops.RemoveUselessCropsPattern'> (enabled = False)
[ INFO ]  Skip replacer <class 'extensions.middle.RemoveDuplicationMemory.RemoveMemoryDuplicationPattern'> (enabled = False)
[ 2019-08-16 11:41:49,895 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.ONNXRNNSequenceNormalize.ONNXRNNSequenceNormalize'>
[ INFO ]  Skip replacer <class 'extensions.middle.TensorIteratorLSTMToLSTMSequence.TensorIteratorLSTM'> (enabled = False)
[ 2019-08-16 11:41:49,909 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.NormalizePad.NormalizePad'>
[ 2019-08-16 11:41:49,926 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.MinimumMiddleReplacer.MinimumMiddleReplacer'>
[ 2019-08-16 11:41:49,942 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.MXNetSplitMultiLayers.MXNetSplitLayersToRNNSequence'>
[ 2019-08-16 11:41:49,957 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.MXNetRNNSequenceNormalize.MXNetRNNSequenceNormalize'>
[ 2019-08-16 11:41:49,972 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.DecomposeBidirectionalRNNSequence.DecomposeBidirectionalRNNSequence'>
[ 2019-08-16 11:41:49,987 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.RNNSequenceNormalizeToIE.RNNSequenceNormalize'>
[ 2019-08-16 11:41:50,001 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.GRURNNSequenceToTensorIterator.GRUAndRNNToTensorIterator'>
[ INFO ]  Skip replacer <class 'extensions.middle.GemmToFullyConnected.GemmToFullyConnected'> (graph_condition not satisfied)
[ 2019-08-16 11:41:50,016 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.wights_permute_normalizer.WeightsPermuteNormalizer'>
[ INFO ]  Skip replacer <class 'extensions.middle.GemmResolver.GemmResolver'> (graph_condition not satisfied)
[ INFO ]  Skip replacer <class 'extensions.middle.NormalizeFullyConnected.NormalizeFullyConnected'> (graph_condition not satisfied)
[ 2019-08-16 11:41:50,042 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.GatherNdNormalizer.GatherNdNormalize'>
[ 2019-08-16 11:41:50,052 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
seq_len/sink_port_0
[ 2019-08-16 11:41:50,062 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
[ 2019-08-16 11:41:50,069 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:50,072 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 130
131
132
133
134
135
136
137
[ 2019-08-16 11:41:50,073 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:50,076 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 130
131
132
133
134
135
136
137
[ 2019-08-16 11:41:50,078 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:50,080 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 130
131
132
133
134
135
136
137
[ 2019-08-16 11:41:50,082 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:50,084 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 130
131
132
133
134
135
136
137
[ 2019-08-16 11:41:50,093 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.FusedBatchNormTrainingCatch.FusedBatchNormTrainingCatch'>
[ 2019-08-16 11:41:50,109 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.FusedBatchNormNonConstant.FusedBatchNormNonConstant'>
[ 2019-08-16 11:41:50,126 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.FlattenONNXToReshape.FlattenONNXToReshape'>
[ 2019-08-16 11:41:50,142 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.EltwiseInputReshape.EltwiseInputReshape'>
[ 2019-08-16 11:41:50,148 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
seq_len/sink_port_0
[ 2019-08-16 11:41:50,159 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
[ 2019-08-16 11:41:50,166 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:50,169 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 138
139
140
141
142
143
144
145
[ 2019-08-16 11:41:50,170 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:50,173 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 138
139
140
141
142
143
144
145
[ 2019-08-16 11:41:50,174 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_1/bidirectional_rnn/bw/bw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:50,177 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 138
139
140
141
142
143
144
145
[ 2019-08-16 11:41:50,178 ] [ DEBUG ] [ eliminate:67 ]  The following nodes are seeded as output reachable:
56/sink_port_0
rnn/stack_bidirectional_rnn/cell_0/bidirectional_rnn/fw/fw/while/gru_cell/add/Output_0/Data_/sink_port_0
[ 2019-08-16 11:41:50,181 ] [ DEBUG ] [ eliminate:133 ]  Removing the following dead nodes: 138
139
140
141
142
143
144
145
[ 2019-08-16 11:41:50,189 ] [ DEBUG ] [ class_registration:267 ]  Run replacer <class 'extensions.middle.EltwiseInputReshape.Eltwise1DInputReshape'>
[ 2019-08-16 11:41:50,190 ] [ DEBUG ] [ op:207 ]  Start running infer function for individual op node with attributes: {'precision': 'FP32', 'kind': 'op', 'type': 'Reshape', 'op': 'Reshape', 'reinterp_shape': True, 'in_ports_count': 2, 'out_ports_count': 1, 'infer': <function Reshape.infer at 0x7fe511aaa048>, 'dim': array([1, 1, 1, 1]), 'name': 'batch_normalization/batchnorm/mul/Output_0/Data_/Broadcast', 'dim_attrs': ['batch_dims', 'channel_dims', 'axis', 'spatial_dims'], 'shape_attrs': ['output_shape', 'pad', 'shape', 'stride', 'window'], 'IE': [('layer', [('id', <function Op.substitute_ie_attrs.<locals>.<lambda> at 0x7fe50a39ebf8>), 'name', 'precision', 'type'], [('data', [], []), '@ports', '@consts'])], '_in_ports': {0: {}, 1: {}}, '_out_ports': {0: {}}}
[ ERROR ]  -------------------------------------------------
[ ERROR ]  ----------------- INTERNAL ERROR ----------------
[ ERROR ]  Unexpected exception happened.
[ ERROR ]  Please contact Model Optimizer developers and forward the following information:
[ ERROR ]  Exception occurred during running replacer "REPLACEMENT_ID (<class 'extensions.middle.EltwiseInputReshape.Eltwise1DInputReshape'>)": The Reshape operation batch_normalization/batchnorm/mul/Output_0/Data_/Broadcast must have 2 inputs
[ ERROR ]  Traceback (most recent call last):
  File "/opt/intel/openvino_2019.2.242/deployment_tools/model_optimizer/mo/utils/class_registration.py", line 273, in apply_replacements
    for_graph_and_each_sub_graph_recursively(graph, replacer.find_and_replace_pattern)
  File "/opt/intel/openvino_2019.2.242/deployment_tools/model_optimizer/mo/middle/pattern_match.py", line 58, in for_graph_and_each_sub_graph_recursively
    func(graph)
  File "/opt/intel/openvino_2019.2.242/deployment_tools/model_optimizer/extensions/middle/EltwiseInputReshape.py", line 67, in find_and_replace_pattern
    reshape_data_node = reshape_data_op.create_node_with_data([node])
  File "/opt/intel/openvino_2019.2.242/deployment_tools/model_optimizer/mo/ops/op.py", line 208, in create_node_with_data
    new_op_node.infer(new_op_node)
  File "/opt/intel/openvino_2019.2.242/deployment_tools/model_optimizer/mo/ops/reshape.py", line 47, in infer
    lambda node: np.reshape(node.in_node(0).value, node.out_node().shape)
  File "/opt/intel/openvino_2019.2.242/deployment_tools/model_optimizer/mo/front/common/partial_infer/elemental.py", line 19, in single_output_infer
    node.out_node(0).shape = shape_infer(node)
  File "/opt/intel/openvino_2019.2.242/deployment_tools/model_optimizer/mo/front/common/partial_infer/reshape.py", line 30, in tf_reshape_shape_infer
    assert len(node.in_nodes()) == 2, 'The Reshape operation {} must have 2 inputs'.format(node.name)
AssertionError: The Reshape operation batch_normalization/batchnorm/mul/Output_0/Data_/Broadcast must have 2 inputs

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File "/opt/intel/openvino_2019.2.242/deployment_tools/model_optimizer/mo/main.py", line 302, in main
    return driver(argv)
  File "/opt/intel/openvino_2019.2.242/deployment_tools/model_optimizer/mo/main.py", line 251, in driver
    is_binary=not argv.input_model_is_text)
  File "/opt/intel/openvino_2019.2.242/deployment_tools/model_optimizer/mo/pipeline/tf.py", line 134, in tf2nx
    class_registration.apply_replacements(graph, class_registration.ClassType.MIDDLE_REPLACER)
  File "/opt/intel/openvino_2019.2.242/deployment_tools/model_optimizer/mo/utils/class_registration.py", line 299, in apply_replacements
    )) from err
Exception: Exception occurred during running replacer "REPLACEMENT_ID (<class 'extensions.middle.EltwiseInputReshape.Eltwise1DInputReshape'>)": The Reshape operation batch_normalization/batchnorm/mul/Output_0/Data_/Broadcast must have 2 inputs

[ ERROR ]  ---------------- END OF BUG REPORT --------------
[ ERROR ]  -------------------------------------------------
Model Optimizer arguments:
Common parameters:
	- Path to the Input Model: 	/opt/intel/openvino_2019.2.242/deployment_tools/model_optimizer/model.pb
	- Path for generated IR: 	/opt/intel/openvino_2019.2.242/deployment_tools/model_optimizer/.
	- IR output name: 	model
	- Log level: 	DEBUG
	- Batch: 	Not specified, inherited from the model
	- Input layers: 	inputs
	- Output layers: 	seq_len
	- Input shapes: 	[1,112,28,1]
	- Mean values: 	Not specified
	- Scale values: 	Not specified
	- Scale factor: 	Not specified
	- Precision of IR: 	FP32
	- Enable fusing: 	True
	- Enable grouped convolutions fusing: 	True
	- Move mean values to preprocess section: 	False
	- Reverse input channels: 	False
TensorFlow specific parameters:
	- Input model in text protobuf format: 	False
	- Path to model dump for TensorBoard: 	None
	- List of shared libraries with TensorFlow custom layers implementation: 	None
	- Update the configuration file with input/output node names: 	None
	- Use configuration file used to generate the model with Object Detection API: 	None
	- Operations to offload: 	None
	- Patterns to offload: 	None
	- Use the config file: 	None
Model Optimizer version: 	2019.2.0-436-gf5827d4