TDA4VM: BEV model deploy

Wu han

Part Number: TDA4VM
Other Parts Discussed in Thread: AM68A

Tool/software:

MatrixVT BEV model deploy demo, this is model、input data and exported info using the script of onnxrt_ep.py with set ‘-c’. But it didn't finish running even after 12 hours.

matrixvt.rar

over 1 year ago

0 Adam Hua over 1 year ago

TI__Expert 5015 points

Hi Wu,

I tried to import this model but get error message as this on 9.2.5:

Can you help set the batch to 1 since batch processing is not currently supported for TDAVM but only supported on VH?

Regards,

Adam

0 Wu han over 1 year ago in reply to Adam Hua

Prodigy 10 points

my sdk version is 9.2.7, it does not has this error, Perhaps my script got stuck before reaching this step of reducemean.
this is the model of the node reducemean with 1 batch.1_batch.rar

0 Adam Hua over 1 year ago in reply to Wu han

TI__Expert 5015 points

Thanks! I will try it.

0 Pratik Kedar over 1 year ago in reply to Adam Hua

TI__Mastermind 24041 points

Adam, better if you can validate on 9.2.7.0

Lets connect internally

0 Adam Hua over 1 year ago in reply to Pratik Kedar

TI__Expert 5015 points

Hi Pratik,

I tried it on 9.2.7 and it still stuck .

python3 multi_input_no_post.py -c -m motorvis_matrix
Available execution providers :  ['TIDLExecutionProvider', 'TIDLCompilationProvider', 'CPUExecutionProvider']

Running 1 Models - ['motorvis_matrix']


Running_Model :  motorvis_matrix  


Running shape inference on model ../../../models/public/MatrixVT_model_2024-05-28-10-33-06_tda4.onnx 

tidl_tools_path                                 = /home/ht/edgeai/edgeai-tidl-tools/tidl_tools 
artifacts_folder                                = ../../../model-artifacts//motorvis_matrix/ 
tidl_tensor_bits                                = 8 
debug_level                                     = 2 
num_tidl_subgraphs                              = 16 
tidl_denylist                                   = 
tidl_denylist_layer_name                        = 
tidl_denylist_layer_type                         = 
tidl_allowlist_layer_name                        = 
model_type                                      =  
tidl_calibration_accuracy_level                 = 7 
tidl_calibration_options:num_frames_calibration = 2 
tidl_calibration_options:bias_calibration_iterations = 5 
mixed_precision_factor = -1.000000 
model_group_id = 0 
power_of_2_quantization                         = 2 
ONNX QDQ Enabled                                = 0 
enable_high_resolution_optimization             = 0 
pre_batchnorm_fold                              = 1 
add_data_convert_ops                            = 3 
output_feature_16bit_names_list                 =  
m_params_16bit_names_list                       =  
m_single_core_layers_names_list                    =  
reserved_compile_constraints_flag               = 1601 
ti_internal_reserved_1                          = 


 ****** WARNING : Network not identified as Object Detection network : (1) Ignore if network is not Object Detection network (2) If network is Object Detection network, please specify "model_type":"OD" as part of OSRT compilation options******

Supported TIDL layer type ---         Reshape -- 1_/Reshape 
Supported TIDL layer type ---            Gemm -- 1_/context_mlp/fc1/Gemm 
Supported TIDL layer type ---            Relu -- 1_/context_mlp/act/Relu 
Supported TIDL layer type ---            Gemm -- 1_/context_mlp/fc2/Gemm 
Supported TIDL layer type ---         Reshape -- 1_/Reshape_1 
Supported TIDL layer type ---            Conv -- 1_/context_se/conv_reduce/Conv 
Supported TIDL layer type ---            Relu -- 1_/context_se/act1/Relu 
Supported TIDL layer type ---            Conv -- 1_/context_se/conv_expand/Conv 
Supported TIDL layer type ---         Sigmoid -- 1_/context_se/gate/Sigmoid 
Supported TIDL layer type ---            Conv -- 0_/0/conv1/Conv 
Supported TIDL layer type ---            Relu -- 0_/0/relu/Relu 
Supported TIDL layer type ---         MaxPool -- 0_/0/maxpool/MaxPool 
Supported TIDL layer type ---            Conv -- 0_/0/layer1/layer1.0/conv1/Conv 
Supported TIDL layer type ---            Relu -- 0_/0/layer1/layer1.0/relu/Relu 
Supported TIDL layer type ---            Conv -- 0_/0/layer1/layer1.0/conv2/Conv 
Supported TIDL layer type ---             Add -- 0_/0/layer1/layer1.0/Add 
Supported TIDL layer type ---            Relu -- 0_/0/layer1/layer1.0/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 0_/0/layer1/layer1.1/conv1/Conv 
Supported TIDL layer type ---            Relu -- 0_/0/layer1/layer1.1/relu/Relu 
Supported TIDL layer type ---            Conv -- 0_/0/layer1/layer1.1/conv2/Conv 
Supported TIDL layer type ---             Add -- 0_/0/layer1/layer1.1/Add 
Supported TIDL layer type ---            Relu -- 0_/0/layer1/layer1.1/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 0_/0/layer2/layer2.0/downsample/downsample.0/Conv 
Supported TIDL layer type ---            Conv -- 0_/0/layer2/layer2.0/conv1/Conv 
Supported TIDL layer type ---            Relu -- 0_/0/layer2/layer2.0/relu/Relu 
Supported TIDL layer type ---            Conv -- 0_/0/layer2/layer2.0/conv2/Conv 
Supported TIDL layer type ---             Add -- 0_/0/layer2/layer2.0/Add 
Supported TIDL layer type ---            Relu -- 0_/0/layer2/layer2.0/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 0_/0/layer2/layer2.1/conv1/Conv 
Supported TIDL layer type ---            Relu -- 0_/0/layer2/layer2.1/relu/Relu 
Supported TIDL layer type ---            Conv -- 0_/0/layer2/layer2.1/conv2/Conv 
Supported TIDL layer type ---             Add -- 0_/0/layer2/layer2.1/Add 
Supported TIDL layer type ---            Relu -- 0_/0/layer2/layer2.1/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 0_/0/layer3/layer3.0/downsample/downsample.0/Conv 
Supported TIDL layer type ---            Conv -- 0_/0/layer3/layer3.0/conv1/Conv 
Supported TIDL layer type ---            Relu -- 0_/0/layer3/layer3.0/relu/Relu 
Supported TIDL layer type ---            Conv -- 0_/0/layer3/layer3.0/conv2/Conv 
Supported TIDL layer type ---             Add -- 0_/0/layer3/layer3.0/Add 
Supported TIDL layer type ---            Relu -- 0_/0/layer3/layer3.0/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 0_/0/layer3/layer3.1/conv1/Conv 
Supported TIDL layer type ---            Relu -- 0_/0/layer3/layer3.1/relu/Relu 
Supported TIDL layer type ---            Conv -- 0_/0/layer3/layer3.1/conv2/Conv 
Supported TIDL layer type ---             Add -- 0_/0/layer3/layer3.1/Add 
Supported TIDL layer type ---            Relu -- 0_/0/layer3/layer3.1/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 0_/1/deblocks.2/deblocks.2.0/Conv 
Supported TIDL layer type ---            Relu -- 0_/1/deblocks.2/deblocks.2.2/Relu 
Supported TIDL layer type ---            Conv -- 0_/1/deblocks.1/deblocks.1.0/Conv 
Supported TIDL layer type ---            Relu -- 0_/1/deblocks.1/deblocks.1.2/Relu 
Supported TIDL layer type ---            Conv -- 0_/1/deblocks.0/deblocks.0.0/Conv 
Supported TIDL layer type ---            Relu -- 0_/1/deblocks.0/deblocks.0.2/Relu 
Supported TIDL layer type ---          Concat -- 0_/1/Concat 
Supported TIDL layer type ---            Conv -- 1_/reduce_conv/reduce_conv.0/Conv 
Supported TIDL layer type ---            Relu -- 1_/reduce_conv/reduce_conv.2/Relu 
Unsupported (TIDL check) TIDL layer type ---             Mul 
Supported TIDL layer type ---            Conv -- 1_/context_conv/Conv 
Supported TIDL layer type ---            Conv -- 2_/vertical_weighter/vertical_weighter.0/Conv 
Supported TIDL layer type ---            Relu -- 2_/vertical_weighter/vertical_weighter.2/Relu 
Supported TIDL layer type ---            Conv -- 2_/vertical_weighter/vertical_weighter.3/Conv 
Supported TIDL layer type ---       Transpose -- 2_/Transpose 
Supported TIDL layer type ---         Softmax -- 2_/Softmax 
Supported TIDL layer type ---       Transpose -- 2_/Transpose_1 
Supported TIDL layer type ---            Conv -- 1_/depth_se/conv_reduce/Conv 
Supported TIDL layer type ---            Relu -- 1_/depth_se/act1/Relu 
Supported TIDL layer type ---            Conv -- 1_/depth_se/conv_expand/Conv 
Supported TIDL layer type ---         Sigmoid -- 1_/depth_se/gate/Sigmoid 
Unsupported (TIDL check) TIDL layer type ---             Mul 
Supported TIDL layer type ---            Conv -- 1_/depth_conv/depth_conv.0/conv1/Conv 
Supported TIDL layer type ---            Relu -- 1_/depth_conv/depth_conv.0/relu/Relu 
Supported TIDL layer type ---            Conv -- 1_/depth_conv/depth_conv.0/conv2/Conv 
Supported TIDL layer type ---             Add -- 1_/depth_conv/depth_conv.0/Add 
Supported TIDL layer type ---            Relu -- 1_/depth_conv/depth_conv.0/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 1_/depth_conv/depth_conv.1/conv1/Conv 
Supported TIDL layer type ---            Relu -- 1_/depth_conv/depth_conv.1/relu/Relu 
Supported TIDL layer type ---            Conv -- 1_/depth_conv/depth_conv.1/conv2/Conv 
Supported TIDL layer type ---             Add -- 1_/depth_conv/depth_conv.1/Add 
Supported TIDL layer type ---            Relu -- 1_/depth_conv/depth_conv.1/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 1_/depth_conv/depth_conv.2/conv1/Conv 
Supported TIDL layer type ---            Relu -- 1_/depth_conv/depth_conv.2/relu/Relu 
Supported TIDL layer type ---            Conv -- 1_/depth_conv/depth_conv.2/conv2/Conv 
Supported TIDL layer type ---             Add -- 1_/depth_conv/depth_conv.2/Add 
Supported TIDL layer type ---            Relu -- 1_/depth_conv/depth_conv.2/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 1_/depth_conv/depth_conv.3/Conv 
Supported TIDL layer type ---         Softmax -- 1_/Softmax 
Unsupported (TIDL check) TIDL layer type ---             Mul 
Supported TIDL layer type ---       Transpose -- 2_/Transpose_2 
Supported TIDL layer type ---         Reshape -- 2_/Reshape 
Supported TIDL layer type ---      ReduceMean -- 2_/ReduceMean 
Supported TIDL layer type ---          MatMul -- 4_/MatMul 
Supported TIDL layer type ---             Mul -- 4_/Mul 
Supported TIDL layer type ---       Transpose -- 4_/Transpose_1 
Supported TIDL layer type ---            Conv -- 3_/context_conv/Conv 
Supported TIDL layer type ---            Conv -- 3_/merger/merger.0/Conv 
Supported TIDL layer type ---         Sigmoid -- 3_/merger/merger.1/Sigmoid 
Supported TIDL layer type ---            Conv -- 3_/merger/merger.2/Conv 
Supported TIDL layer type ---            Conv -- 3_/reduce_conv/reduce_conv.0/Conv 
Supported TIDL layer type ---            Relu -- 3_/reduce_conv/reduce_conv.2/Relu 
Supported TIDL layer type ---            Conv -- 3_/conv1/conv1.0/Conv 
Supported TIDL layer type ---            Relu -- 3_/conv1/conv1.2/Relu 
Supported TIDL layer type ---            Conv -- 3_/conv1/conv1.3/Conv 
Supported TIDL layer type ---            Relu -- 3_/conv1/conv1.5/Relu 
Supported TIDL layer type ---             Add -- 3_/Add 
Supported TIDL layer type ---            Conv -- 3_/conv2/conv2.0/Conv 
Supported TIDL layer type ---            Relu -- 3_/conv2/conv2.2/Relu 
Supported TIDL layer type ---            Conv -- 3_/conv2/conv2.3/Conv 
Supported TIDL layer type ---            Relu -- 3_/conv2/conv2.5/Relu 
Supported TIDL layer type ---             Add -- 3_/Add_1 
Supported TIDL layer type ---            Conv -- 3_/out_conv/out_conv.0/Conv 
Supported TIDL layer type ---            Relu -- 3_/out_conv/out_conv.2/Relu 
Supported TIDL layer type ---       Transpose -- 3_/Transpose 
Supported TIDL layer type ---            Conv -- 3_/gather_conv/Conv 
Supported TIDL layer type ---         Reshape -- 3_/Reshape 
Supported TIDL layer type ---       Transpose -- 4_/Transpose 
Supported TIDL layer type ---         Reshape -- 4_/Reshape 
Supported TIDL layer type ---          MatMul -- 4_/MatMul_1 
Supported TIDL layer type ---       Transpose -- 4_/Transpose_2 
Supported TIDL layer type ---         Reshape -- 4_/Reshape_1 
Supported TIDL layer type ---            Conv -- 5_/conv1/Conv 
Supported TIDL layer type ---            Relu -- 5_/relu/Relu 
Supported TIDL layer type ---            Conv -- 5_/layer1/layer1.0/conv1/Conv 
Supported TIDL layer type ---            Relu -- 5_/layer1/layer1.0/relu/Relu 
Supported TIDL layer type ---            Conv -- 5_/layer1/layer1.0/conv2/Conv 
Supported TIDL layer type ---             Add -- 5_/layer1/layer1.0/Add 
Supported TIDL layer type ---            Relu -- 5_/layer1/layer1.0/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 5_/layer1/layer1.1/conv1/Conv 
Supported TIDL layer type ---            Relu -- 5_/layer1/layer1.1/relu/Relu 
Supported TIDL layer type ---            Conv -- 5_/layer1/layer1.1/conv2/Conv 
Supported TIDL layer type ---             Add -- 5_/layer1/layer1.1/Add 
Supported TIDL layer type ---            Relu -- 5_/layer1/layer1.1/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 5_/layer2/layer2.0/downsample/downsample.0/Conv 
Supported TIDL layer type ---            Conv -- 5_/layer2/layer2.0/conv1/Conv 
Supported TIDL layer type ---            Relu -- 5_/layer2/layer2.0/relu/Relu 
Supported TIDL layer type ---            Conv -- 5_/layer2/layer2.0/conv2/Conv 
Supported TIDL layer type ---             Add -- 5_/layer2/layer2.0/Add 
Supported TIDL layer type ---            Relu -- 5_/layer2/layer2.0/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 5_/layer2/layer2.1/conv1/Conv 
Supported TIDL layer type ---            Relu -- 5_/layer2/layer2.1/relu/Relu 
Supported TIDL layer type ---            Conv -- 5_/layer2/layer2.1/conv2/Conv 
Supported TIDL layer type ---             Add -- 5_/layer2/layer2.1/Add 
Supported TIDL layer type ---            Relu -- 5_/layer2/layer2.1/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 5_/layer3/layer3.0/downsample/downsample.0/Conv 
Supported TIDL layer type ---            Conv -- 5_/layer3/layer3.0/conv1/Conv 
Supported TIDL layer type ---            Relu -- 5_/layer3/layer3.0/relu/Relu 
Supported TIDL layer type ---            Conv -- 5_/layer3/layer3.0/conv2/Conv 
Supported TIDL layer type ---             Add -- 5_/layer3/layer3.0/Add 
Supported TIDL layer type ---            Relu -- 5_/layer3/layer3.0/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 5_/layer3/layer3.1/conv1/Conv 
Supported TIDL layer type ---            Relu -- 5_/layer3/layer3.1/relu/Relu 
Supported TIDL layer type ---            Conv -- 5_/layer3/layer3.1/conv2/Conv 
Supported TIDL layer type ---             Add -- 5_/layer3/layer3.1/Add 
Supported TIDL layer type ---            Relu -- 5_/layer3/layer3.1/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 5_/neck/deblocks.3/deblocks.3.0/Conv 
Supported TIDL layer type ---            Relu -- 5_/neck/deblocks.3/deblocks.3.2/Relu 
Supported TIDL layer type ---          Resize -- 5_/neck/Resize_2 
Supported TIDL layer type ---            Conv -- 5_/neck/deblocks.2/deblocks.2.0/Conv 
Supported TIDL layer type ---            Relu -- 5_/neck/deblocks.2/deblocks.2.2/Relu 
Supported TIDL layer type ---          Resize -- 5_/neck/Resize_1 
Supported TIDL layer type ---            Conv -- 5_/neck/deblocks.1/deblocks.1.0/Conv 
Supported TIDL layer type ---            Relu -- 5_/neck/deblocks.1/deblocks.1.2/Relu 
Supported TIDL layer type ---          Resize -- 5_/neck/Resize 
Supported TIDL layer type ---            Conv -- 5_/neck/deblocks.0/deblocks.0.0/Conv 
Supported TIDL layer type ---            Relu -- 5_/neck/deblocks.0/deblocks.0.2/Relu 
Supported TIDL layer type ---          Concat -- 5_/neck/Concat 
Supported TIDL layer type ---            Conv -- 5_/shared_conv/conv/Conv 
Supported TIDL layer type ---            Relu -- 5_/shared_conv/activate/Relu 
Supported TIDL layer type ---            Conv -- 5_/task_heads.0/heatmap/heatmap.0/conv/Conv 
Supported TIDL layer type ---            Relu -- 5_/task_heads.0/heatmap/heatmap.0/activate/Relu 
Supported TIDL layer type ---            Conv -- 5_/task_heads.0/heatmap/heatmap.1/Conv 
Supported TIDL layer type ---            Conv -- 5_/task_heads.0/vel/vel.0/conv/Conv 
Supported TIDL layer type ---            Relu -- 5_/task_heads.0/vel/vel.0/activate/Relu 
Supported TIDL layer type ---            Conv -- 5_/task_heads.0/vel/vel.1/Conv 
Supported TIDL layer type ---            Conv -- 5_/task_heads.0/rot/rot.0/conv/Conv 
Supported TIDL layer type ---            Relu -- 5_/task_heads.0/rot/rot.0/activate/Relu 
Supported TIDL layer type ---            Conv -- 5_/task_heads.0/rot/rot.1/Conv 
Supported TIDL layer type ---            Conv -- 5_/task_heads.0/dim/dim.0/conv/Conv 
Supported TIDL layer type ---            Relu -- 5_/task_heads.0/dim/dim.0/activate/Relu 
Supported TIDL layer type ---            Conv -- 5_/task_heads.0/dim/dim.1/Conv 
Supported TIDL layer type ---            Conv -- 5_/task_heads.0/height/height.0/conv/Conv 
Supported TIDL layer type ---            Relu -- 5_/task_heads.0/height/height.0/activate/Relu 
Supported TIDL layer type ---            Conv -- 5_/task_heads.0/height/height.1/Conv 
Supported TIDL layer type ---            Conv -- 5_/task_heads.0/reg/reg.0/conv/Conv 
Supported TIDL layer type ---            Relu -- 5_/task_heads.0/reg/reg.0/activate/Relu 
Supported TIDL layer type ---            Conv -- 5_/task_heads.0/reg/reg.1/Conv 

Preliminary subgraphs created = 4 
Final number of subgraphs created are : 4, - Offloaded Nodes - 179, Total Nodes - 182 
ALLOWLISTING : ADD/MUL/SUB/DIV layer : The variable inputs must of be same dimensions --tidl_import_common_model_check.cpp , TIDL_checkAddMulSubDivTensorProperties , 403  
ALLOWLISTING : ADD/MUL/SUB/DIV layer : The variable inputs must of be same dimensions --tidl_import_common_model_check.cpp , TIDL_checkAddMulSubDivTensorProperties , 403  
ALLOWLISTING : ADD/MUL/SUB/DIV layer : The variable inputs must of be same dimensions --tidl_import_common_model_check.cpp , TIDL_checkAddMulSubDivTensorProperties , 403  
SUGGESTION -- [TIDL_ResizeLayer]  Resize kernel with non-power of 2 resize ratio is not optimal.  
INFORMATION -- [TIDL_ResizeLayer]  Any resize ratio which is power of 2 and greater than 4 will be placed by combination of 4x4 resize layer and 2x2 resize layer. For example a 8x8 resize will be replaced by 4x4 resize followed by 2x2 resize.  
INFORMATION -- [TIDL_ResizeLayer]  Any resize ratio which is power of 2 and greater than 4 will be placed by combination of 4x4 resize layer and 2x2 resize layer. For example a 8x8 resize will be replaced by 4x4 resize followed by 2x2 resize.  
Running runtimes graphviz - /home/ht/edgeai/edgeai-tidl-tools/tidl_tools/tidl_graphVisualiser_runtimes.out ../../../model-artifacts//motorvis_matrix//allowedNode.txt ../../../model-artifacts//motorvis_matrix//tempDir/graphvizInfo.txt ../../../model-artifacts//motorvis_matrix//tempDir/runtimes_visualization.svg 
*** In TIDL_createStateImportFunc *** 
Compute on node : TIDLExecutionProvider_TIDL_0_0
  0,            Conv, 3, 1, 0_img, 0_/0/conv1/Conv_output_0
  1,            Relu, 1, 1, 0_/0/conv1/Conv_output_0, 0_/0/relu/Relu_output_0
  2,         MaxPool, 1, 1, 0_/0/relu/Relu_output_0, 0_/0/maxpool/MaxPool_output_0
  3,            Conv, 3, 1, 0_/0/maxpool/MaxPool_output_0, 0_/0/layer1/layer1.0/conv1/Conv_output_0
  4,            Relu, 1, 1, 0_/0/layer1/layer1.0/conv1/Conv_output_0, 0_/0/layer1/layer1.0/relu/Relu_output_0
  5,            Conv, 3, 1, 0_/0/layer1/layer1.0/relu/Relu_output_0, 0_/0/layer1/layer1.0/conv2/Conv_output_0
  6,             Add, 2, 1, 0_/0/layer1/layer1.0/conv2/Conv_output_0, 0_/0/layer1/layer1.0/Add_output_0
  7,            Relu, 1, 1, 0_/0/layer1/layer1.0/Add_output_0, 0_/0/layer1/layer1.0/relu_1/Relu_output_0
  8,            Conv, 3, 1, 0_/0/layer1/layer1.0/relu_1/Relu_output_0, 0_/0/layer1/layer1.1/conv1/Conv_output_0
  9,            Relu, 1, 1, 0_/0/layer1/layer1.1/conv1/Conv_output_0, 0_/0/layer1/layer1.1/relu/Relu_output_0
 10,            Conv, 3, 1, 0_/0/layer1/layer1.1/relu/Relu_output_0, 0_/0/layer1/layer1.1/conv2/Conv_output_0
 11,             Add, 2, 1, 0_/0/layer1/layer1.1/conv2/Conv_output_0, 0_/0/layer1/layer1.1/Add_output_0
 12,            Relu, 1, 1, 0_/0/layer1/layer1.1/Add_output_0, 0_/0/layer1/layer1.1/relu_1/Relu_output_0
 13,            Conv, 3, 1, 0_/0/layer1/layer1.1/relu_1/Relu_output_0, 0_/1/deblocks.0/deblocks.0.0/Conv_output_0
 14,            Relu, 1, 1, 0_/1/deblocks.0/deblocks.0.0/Conv_output_0, 0_/1/deblocks.0/deblocks.0.2/Relu_output_0
 15,            Conv, 3, 1, 0_/0/layer1/layer1.1/relu_1/Relu_output_0, 0_/0/layer2/layer2.0/conv1/Conv_output_0
 16,            Relu, 1, 1, 0_/0/layer2/layer2.0/conv1/Conv_output_0, 0_/0/layer2/layer2.0/relu/Relu_output_0
 17,            Conv, 3, 1, 0_/0/layer2/layer2.0/relu/Relu_output_0, 0_/0/layer2/layer2.0/conv2/Conv_output_0
 18,            Conv, 3, 1, 0_/0/layer1/layer1.1/relu_1/Relu_output_0, 0_/0/layer2/layer2.0/downsample/downsample.0/Conv_output_0
 19,             Add, 2, 1, 0_/0/layer2/layer2.0/conv2/Conv_output_0, 0_/0/layer2/layer2.0/Add_output_0
 20,            Relu, 1, 1, 0_/0/layer2/layer2.0/Add_output_0, 0_/0/layer2/layer2.0/relu_1/Relu_output_0
 21,            Conv, 3, 1, 0_/0/layer2/layer2.0/relu_1/Relu_output_0, 0_/0/layer2/layer2.1/conv1/Conv_output_0
 22,            Relu, 1, 1, 0_/0/layer2/layer2.1/conv1/Conv_output_0, 0_/0/layer2/layer2.1/relu/Relu_output_0
 23,            Conv, 3, 1, 0_/0/layer2/layer2.1/relu/Relu_output_0, 0_/0/layer2/layer2.1/conv2/Conv_output_0
 24,             Add, 2, 1, 0_/0/layer2/layer2.1/conv2/Conv_output_0, 0_/0/layer2/layer2.1/Add_output_0
 25,            Relu, 1, 1, 0_/0/layer2/layer2.1/Add_output_0, 0_/0/layer2/layer2.1/relu_1/Relu_output_0
 26,            Conv, 3, 1, 0_/0/layer2/layer2.1/relu_1/Relu_output_0, 0_/1/deblocks.1/deblocks.1.0/Conv_output_0
 27,            Relu, 1, 1, 0_/1/deblocks.1/deblocks.1.0/Conv_output_0, 0_/1/deblocks.1/deblocks.1.2/Relu_output_0
 28,            Conv, 3, 1, 0_/0/layer2/layer2.1/relu_1/Relu_output_0, 0_/0/layer3/layer3.0/conv1/Conv_output_0
 29,            Relu, 1, 1, 0_/0/layer3/layer3.0/conv1/Conv_output_0, 0_/0/layer3/layer3.0/relu/Relu_output_0
 30,            Conv, 3, 1, 0_/0/layer3/layer3.0/relu/Relu_output_0, 0_/0/layer3/layer3.0/conv2/Conv_output_0
 31,            Conv, 3, 1, 0_/0/layer2/layer2.1/relu_1/Relu_output_0, 0_/0/layer3/layer3.0/downsample/downsample.0/Conv_output_0
 32,             Add, 2, 1, 0_/0/layer3/layer3.0/conv2/Conv_output_0, 0_/0/layer3/layer3.0/Add_output_0
 33,            Relu, 1, 1, 0_/0/layer3/layer3.0/Add_output_0, 0_/0/layer3/layer3.0/relu_1/Relu_output_0
 34,            Conv, 3, 1, 0_/0/layer3/layer3.0/relu_1/Relu_output_0, 0_/0/layer3/layer3.1/conv1/Conv_output_0
 35,            Relu, 1, 1, 0_/0/layer3/layer3.1/conv1/Conv_output_0, 0_/0/layer3/layer3.1/relu/Relu_output_0
 36,            Conv, 3, 1, 0_/0/layer3/layer3.1/relu/Relu_output_0, 0_/0/layer3/layer3.1/conv2/Conv_output_0
 37,             Add, 2, 1, 0_/0/layer3/layer3.1/conv2/Conv_output_0, 0_/0/layer3/layer3.1/Add_output_0
 38,            Relu, 1, 1, 0_/0/layer3/layer3.1/Add_output_0, 0_/0/layer3/layer3.1/relu_1/Relu_output_0
 39,            Conv, 3, 1, 0_/0/layer3/layer3.1/relu_1/Relu_output_0, 0_/1/deblocks.2/deblocks.2.0/Conv_output_0
 40,            Relu, 1, 1, 0_/1/deblocks.2/deblocks.2.0/Conv_output_0, 0_/1/deblocks.2/deblocks.2.2/Relu_output_0
 41,          Concat, 3, 1, 0_/1/deblocks.0/deblocks.0.2/Relu_output_0, 0_backbone_feature
 42,            Conv, 3, 1, 0_backbone_feature, 1_/reduce_conv/reduce_conv.0/Conv_output_0
 43,            Relu, 1, 1, 1_/reduce_conv/reduce_conv.0/Conv_output_0, 1_/reduce_conv/reduce_conv.2/Relu_output_0
 44,         Reshape, 2, 1, 1_mlp_input, 1_/Reshape_output_0
 45,            Gemm, 3, 1, 1_/Reshape_output_0, 1_/context_mlp/fc1/Gemm_output_0
 46,            Relu, 1, 1, 1_/context_mlp/fc1/Gemm_output_0, 1_/context_mlp/act/Relu_output_0
 47,            Gemm, 3, 1, 1_/context_mlp/act/Relu_output_0, 1_/context_mlp/fc2/Gemm_output_0
 48,         Reshape, 2, 1, 1_/context_mlp/fc2/Gemm_output_0, 1_/Reshape_1_output_0
 49,            Conv, 3, 1, 1_/Reshape_1_output_0, 1_/context_se/conv_reduce/Conv_output_0
 50,            Relu, 1, 1, 1_/context_se/conv_reduce/Conv_output_0, 1_/context_se/act1/Relu_output_0
 51,            Conv, 3, 1, 1_/context_se/act1/Relu_output_0, 1_/context_se/conv_expand/Conv_output_0
 52,         Sigmoid, 1, 1, 1_/context_se/conv_expand/Conv_output_0, 1_/context_se/gate/Sigmoid_output_0

Input tensor name -  1_mlp_input 

Input tensor name -  0_img 
Output tensor name - 1_/Reshape_1_output_0 
Output tensor name - 1_/context_se/gate/Sigmoid_output_0 
Output tensor name - 1_/reduce_conv/reduce_conv.2/Relu_output_0 
*** In TIDL_createStateImportFunc *** 
Compute on node : TIDLExecutionProvider_TIDL_1_1
  0,            Conv, 3, 1, 1_/Reshape_1_output_0, 1_/depth_se/conv_reduce/Conv_output_0
  1,            Relu, 1, 1, 1_/depth_se/conv_reduce/Conv_output_0, 1_/depth_se/act1/Relu_output_0
  2,            Conv, 3, 1, 1_/depth_se/act1/Relu_output_0, 1_/depth_se/conv_expand/Conv_output_0
  3,         Sigmoid, 1, 1, 1_/depth_se/conv_expand/Conv_output_0, 1_/depth_se/gate/Sigmoid_output_0
  4,            Conv, 3, 1, 1_/context_se/Mul_output_0, 1_context
  5,            Conv, 3, 1, 1_context, 2_/vertical_weighter/vertical_weighter.0/Conv_output_0
  6,            Relu, 1, 1, 2_/vertical_weighter/vertical_weighter.0/Conv_output_0, 2_/vertical_weighter/vertical_weighter.2/Relu_output_0
  7,            Conv, 3, 1, 2_/vertical_weighter/vertical_weighter.2/Relu_output_0, 2_/vertical_weighter/vertical_weighter.3/Conv_output_0
  8,       Transpose, 1, 1, 2_/vertical_weighter/vertical_weighter.3/Conv_output_0, 2_/Transpose_output_0
  9,         Softmax, 1, 1, 2_/Transpose_output_0, 2_/Softmax_output_0
 10,       Transpose, 1, 1, 2_/Softmax_output_0, 2_/Transpose_1_output_0

Input tensor name -  1_/context_se/Mul_output_0 

Input tensor name -  1_/Reshape_1_output_0 
Output tensor name - 1_context 
Output tensor name - 2_/Transpose_1_output_0 
Output tensor name - 1_/depth_se/gate/Sigmoid_output_0 
*** In TIDL_createStateImportFunc *** 
Compute on node : TIDLExecutionProvider_TIDL_2_2
  0,            Conv, 3, 1, 1_/depth_se/Mul_output_0, 1_/depth_conv/depth_conv.0/conv1/Conv_output_0
  1,            Relu, 1, 1, 1_/depth_conv/depth_conv.0/conv1/Conv_output_0, 1_/depth_conv/depth_conv.0/relu/Relu_output_0
  2,            Conv, 3, 1, 1_/depth_conv/depth_conv.0/relu/Relu_output_0, 1_/depth_conv/depth_conv.0/conv2/Conv_output_0
  3,             Add, 2, 1, 1_/depth_conv/depth_conv.0/conv2/Conv_output_0, 1_/depth_conv/depth_conv.0/Add_output_0
  4,            Relu, 1, 1, 1_/depth_conv/depth_conv.0/Add_output_0, 1_/depth_conv/depth_conv.0/relu_1/Relu_output_0
  5,            Conv, 3, 1, 1_/depth_conv/depth_conv.0/relu_1/Relu_output_0, 1_/depth_conv/depth_conv.1/conv1/Conv_output_0
  6,            Relu, 1, 1, 1_/depth_conv/depth_conv.1/conv1/Conv_output_0, 1_/depth_conv/depth_conv.1/relu/Relu_output_0
  7,            Conv, 3, 1, 1_/depth_conv/depth_conv.1/relu/Relu_output_0, 1_/depth_conv/depth_conv.1/conv2/Conv_output_0
  8,             Add, 2, 1, 1_/depth_conv/depth_conv.1/conv2/Conv_output_0, 1_/depth_conv/depth_conv.1/Add_output_0
  9,            Relu, 1, 1, 1_/depth_conv/depth_conv.1/Add_output_0, 1_/depth_conv/depth_conv.1/relu_1/Relu_output_0
 10,            Conv, 3, 1, 1_/depth_conv/depth_conv.1/relu_1/Relu_output_0, 1_/depth_conv/depth_conv.2/conv1/Conv_output_0
 11,            Relu, 1, 1, 1_/depth_conv/depth_conv.2/conv1/Conv_output_0, 1_/depth_conv/depth_conv.2/relu/Relu_output_0
 12,            Conv, 3, 1, 1_/depth_conv/depth_conv.2/relu/Relu_output_0, 1_/depth_conv/depth_conv.2/conv2/Conv_output_0
 13,             Add, 2, 1, 1_/depth_conv/depth_conv.2/conv2/Conv_output_0, 1_/depth_conv/depth_conv.2/Add_output_0
 14,            Relu, 1, 1, 1_/depth_conv/depth_conv.2/Add_output_0, 1_/depth_conv/depth_conv.2/relu_1/Relu_output_0
 15,            Conv, 3, 1, 1_/depth_conv/depth_conv.2/relu_1/Relu_output_0, 1_/depth_conv/depth_conv.3/Conv_output_0
 16,         Softmax, 1, 1, 1_/depth_conv/depth_conv.3/Conv_output_0, 1_depth

Input tensor name -  1_/depth_se/Mul_output_0 
Output tensor name - 1_depth 
*** In TIDL_createStateImportFunc *** 
Compute on node : TIDLExecutionProvider_TIDL_3_3
  0,            Conv, 3, 1, 1_context, 3_/context_conv/Conv_output_0
  1,            Conv, 3, 1, 3_/context_conv/Conv_output_0, 3_/merger/merger.0/Conv_output_0
  2,         Sigmoid, 1, 1, 3_/merger/merger.0/Conv_output_0, 3_/merger/merger.1/Sigmoid_output_0
  3,            Conv, 3, 1, 3_/merger/merger.1/Sigmoid_output_0, 3_/merger/merger.2/Conv_output_0
  4,            Conv, 3, 1, 3_/merger/merger.2/Conv_output_0, 3_/reduce_conv/reduce_conv.0/Conv_output_0
  5,            Relu, 1, 1, 3_/reduce_conv/reduce_conv.0/Conv_output_0, 3_/reduce_conv/reduce_conv.2/Relu_output_0
  6,            Conv, 3, 1, 3_/reduce_conv/reduce_conv.2/Relu_output_0, 3_/conv1/conv1.0/Conv_output_0
  7,            Relu, 1, 1, 3_/conv1/conv1.0/Conv_output_0, 3_/conv1/conv1.2/Relu_output_0
  8,            Conv, 3, 1, 3_/conv1/conv1.2/Relu_output_0, 3_/conv1/conv1.3/Conv_output_0
  9,            Relu, 1, 1, 3_/conv1/conv1.3/Conv_output_0, 3_/conv1/conv1.5/Relu_output_0
 10,             Add, 2, 1, 3_/conv1/conv1.5/Relu_output_0, 3_/Add_output_0
 11,            Conv, 3, 1, 3_/Add_output_0, 3_/conv2/conv2.0/Conv_output_0
 12,            Relu, 1, 1, 3_/conv2/conv2.0/Conv_output_0, 3_/conv2/conv2.2/Relu_output_0
 13,            Conv, 3, 1, 3_/conv2/conv2.2/Relu_output_0, 3_/conv2/conv2.3/Conv_output_0
 14,            Relu, 1, 1, 3_/conv2/conv2.3/Conv_output_0, 3_/conv2/conv2.5/Relu_output_0
 15,             Add, 2, 1, 3_/conv2/conv2.5/Relu_output_0, 3_/Add_1_output_0
 16,            Conv, 3, 1, 3_/Add_1_output_0, 3_/out_conv/out_conv.0/Conv_output_0
 17,            Relu, 1, 1, 3_/out_conv/out_conv.0/Conv_output_0, 3_/out_conv/out_conv.2/Relu_output_0
 18,       Transpose, 1, 1, 3_/out_conv/out_conv.2/Relu_output_0, 3_/Transpose_output_0
 19,            Conv, 3, 1, 3_/Transpose_output_0, 3_/gather_conv/Conv_output_0
 20,         Reshape, 2, 1, 3_/gather_conv/Conv_output_0, 3_HoriConv_out
 21,       Transpose, 1, 1, 3_HoriConv_out, 4_/Transpose_output_0
 22,         Reshape, 2, 1, 4_/Transpose_output_0, 4_/Reshape_output_0
 23,       Transpose, 1, 1, 2_/Mul_output_0, 2_/Transpose_2_output_0
 24,         Reshape, 2, 1, 2_/Transpose_2_output_0, 2_/Reshape_output_0
 25,      ReduceMean, 1, 1, 2_/Reshape_output_0, 2_DepthReducer_out
 26,          MatMul, 2, 1, 2_DepthReducer_out, 4_/MatMul_output_0
 27,             Mul, 2, 1, 4_/MatMul_output_0, 4_/Mul_output_0
 28,       Transpose, 1, 1, 4_/Mul_output_0, 4_/Transpose_1_output_0
 29,          MatMul, 2, 1, 4_/Transpose_1_output_0, 4_/MatMul_1_output_0
 30,       Transpose, 1, 1, 4_/MatMul_1_output_0, 4_/Transpose_2_output_0
 31,         Reshape, 2, 1, 4_/Transpose_2_output_0, 4_bev_out
 32,            Conv, 3, 1, 4_bev_out, 5_/neck/deblocks.0/deblocks.0.0/Conv_output_0
 33,            Relu, 1, 1, 5_/neck/deblocks.0/deblocks.0.0/Conv_output_0, 5_/neck/deblocks.0/deblocks.0.2/Relu_output_0
 34,            Conv, 3, 1, 4_bev_out, 5_/conv1/Conv_output_0
 35,            Relu, 1, 1, 5_/conv1/Conv_output_0, 5_/relu/Relu_output_0
 36,            Conv, 3, 1, 5_/relu/Relu_output_0, 5_/layer1/layer1.0/conv1/Conv_output_0
 37,            Relu, 1, 1, 5_/layer1/layer1.0/conv1/Conv_output_0, 5_/layer1/layer1.0/relu/Relu_output_0
 38,            Conv, 3, 1, 5_/layer1/layer1.0/relu/Relu_output_0, 5_/layer1/layer1.0/conv2/Conv_output_0
 39,             Add, 2, 1, 5_/layer1/layer1.0/conv2/Conv_output_0, 5_/layer1/layer1.0/Add_output_0
 40,            Relu, 1, 1, 5_/layer1/layer1.0/Add_output_0, 5_/layer1/layer1.0/relu_1/Relu_output_0
 41,            Conv, 3, 1, 5_/layer1/layer1.0/relu_1/Relu_output_0, 5_/layer1/layer1.1/conv1/Conv_output_0
 42,            Relu, 1, 1, 5_/layer1/layer1.1/conv1/Conv_output_0, 5_/layer1/layer1.1/relu/Relu_output_0
 43,            Conv, 3, 1, 5_/layer1/layer1.1/relu/Relu_output_0, 5_/layer1/layer1.1/conv2/Conv_output_0
 44,             Add, 2, 1, 5_/layer1/layer1.1/conv2/Conv_output_0, 5_/layer1/layer1.1/Add_output_0
 45,            Relu, 1, 1, 5_/layer1/layer1.1/Add_output_0, 5_/layer1/layer1.1/relu_1/Relu_output_0
 46,            Conv, 3, 1, 5_/layer1/layer1.1/relu_1/Relu_output_0, 5_/neck/deblocks.1/deblocks.1.0/Conv_output_0
 47,            Relu, 1, 1, 5_/neck/deblocks.1/deblocks.1.0/Conv_output_0, 5_/neck/deblocks.1/deblocks.1.2/Relu_output_0
 48,          Resize, 3, 1, 5_/neck/deblocks.1/deblocks.1.2/Relu_output_0, 5_/neck/Resize_output_0
 49,            Conv, 3, 1, 5_/layer1/layer1.1/relu_1/Relu_output_0, 5_/layer2/layer2.0/conv1/Conv_output_0
 50,            Relu, 1, 1, 5_/layer2/layer2.0/conv1/Conv_output_0, 5_/layer2/layer2.0/relu/Relu_output_0
 51,            Conv, 3, 1, 5_/layer2/layer2.0/relu/Relu_output_0, 5_/layer2/layer2.0/conv2/Conv_output_0
 52,            Conv, 3, 1, 5_/layer1/layer1.1/relu_1/Relu_output_0, 5_/layer2/layer2.0/downsample/downsample.0/Conv_output_0
 53,             Add, 2, 1, 5_/layer2/layer2.0/conv2/Conv_output_0, 5_/layer2/layer2.0/Add_output_0
 54,            Relu, 1, 1, 5_/layer2/layer2.0/Add_output_0, 5_/layer2/layer2.0/relu_1/Relu_output_0
 55,            Conv, 3, 1, 5_/layer2/layer2.0/relu_1/Relu_output_0, 5_/layer2/layer2.1/conv1/Conv_output_0
 56,            Relu, 1, 1, 5_/layer2/layer2.1/conv1/Conv_output_0, 5_/layer2/layer2.1/relu/Relu_output_0
 57,            Conv, 3, 1, 5_/layer2/layer2.1/relu/Relu_output_0, 5_/layer2/layer2.1/conv2/Conv_output_0
 58,             Add, 2, 1, 5_/layer2/layer2.1/conv2/Conv_output_0, 5_/layer2/layer2.1/Add_output_0
 59,            Relu, 1, 1, 5_/layer2/layer2.1/Add_output_0, 5_/layer2/layer2.1/relu_1/Relu_output_0
 60,            Conv, 3, 1, 5_/layer2/layer2.1/relu_1/Relu_output_0, 5_/neck/deblocks.2/deblocks.2.0/Conv_output_0
 61,            Relu, 1, 1, 5_/neck/deblocks.2/deblocks.2.0/Conv_output_0, 5_/neck/deblocks.2/deblocks.2.2/Relu_output_0
 62,          Resize, 3, 1, 5_/neck/deblocks.2/deblocks.2.2/Relu_output_0, 5_/neck/Resize_1_output_0
 63,            Conv, 3, 1, 5_/layer2/layer2.1/relu_1/Relu_output_0, 5_/layer3/layer3.0/conv1/Conv_output_0
 64,            Relu, 1, 1, 5_/layer3/layer3.0/conv1/Conv_output_0, 5_/layer3/layer3.0/relu/Relu_output_0
 65,            Conv, 3, 1, 5_/layer3/layer3.0/relu/Relu_output_0, 5_/layer3/layer3.0/conv2/Conv_output_0
 66,            Conv, 3, 1, 5_/layer2/layer2.1/relu_1/Relu_output_0, 5_/layer3/layer3.0/downsample/downsample.0/Conv_output_0
 67,             Add, 2, 1, 5_/layer3/layer3.0/conv2/Conv_output_0, 5_/layer3/layer3.0/Add_output_0
 68,            Relu, 1, 1, 5_/layer3/layer3.0/Add_output_0, 5_/layer3/layer3.0/relu_1/Relu_output_0
 69,            Conv, 3, 1, 5_/layer3/layer3.0/relu_1/Relu_output_0, 5_/layer3/layer3.1/conv1/Conv_output_0
 70,            Relu, 1, 1, 5_/layer3/layer3.1/conv1/Conv_output_0, 5_/layer3/layer3.1/relu/Relu_output_0
 71,            Conv, 3, 1, 5_/layer3/layer3.1/relu/Relu_output_0, 5_/layer3/layer3.1/conv2/Conv_output_0
 72,             Add, 2, 1, 5_/layer3/layer3.1/conv2/Conv_output_0, 5_/layer3/layer3.1/Add_output_0
 73,            Relu, 1, 1, 5_/layer3/layer3.1/Add_output_0, 5_/layer3/layer3.1/relu_1/Relu_output_0
 74,            Conv, 3, 1, 5_/layer3/layer3.1/relu_1/Relu_output_0, 5_/neck/deblocks.3/deblocks.3.0/Conv_output_0
 75,            Relu, 1, 1, 5_/neck/deblocks.3/deblocks.3.0/Conv_output_0, 5_/neck/deblocks.3/deblocks.3.2/Relu_output_0
 76,          Resize, 3, 1, 5_/neck/deblocks.3/deblocks.3.2/Relu_output_0, 5_/neck/Resize_2_output_0
 77,          Concat, 4, 1, 5_/neck/deblocks.0/deblocks.0.2/Relu_output_0, 5_/neck/Concat_output_0
 78,            Conv, 3, 1, 5_/neck/Concat_output_0, 5_/shared_conv/conv/Conv_output_0
 79,            Relu, 1, 1, 5_/shared_conv/conv/Conv_output_0, 5_/shared_conv/activate/Relu_output_0
 80,            Conv, 3, 1, 5_/shared_conv/activate/Relu_output_0, 5_/task_heads.0/reg/reg.0/conv/Conv_output_0
 81,            Relu, 1, 1, 5_/task_heads.0/reg/reg.0/conv/Conv_output_0, 5_/task_heads.0/reg/reg.0/activate/Relu_output_0
 82,            Conv, 3, 1, 5_/task_heads.0/reg/reg.0/activate/Relu_output_0, 5_preds
 83,            Conv, 3, 1, 5_/shared_conv/activate/Relu_output_0, 5_/task_heads.0/height/height.0/conv/Conv_output_0
 84,            Relu, 1, 1, 5_/task_heads.0/height/height.0/conv/Conv_output_0, 5_/task_heads.0/height/height.0/activate/Relu_output_0
 85,            Conv, 3, 1, 5_/task_heads.0/height/height.0/activate/Relu_output_0, 5_257
 86,            Conv, 3, 1, 5_/shared_conv/activate/Relu_output_0, 5_/task_heads.0/dim/dim.0/conv/Conv_output_0
 87,            Relu, 1, 1, 5_/task_heads.0/dim/dim.0/conv/Conv_output_0, 5_/task_heads.0/dim/dim.0/activate/Relu_output_0
 88,            Conv, 3, 1, 5_/task_heads.0/dim/dim.0/activate/Relu_output_0, 5_261
 89,            Conv, 3, 1, 5_/shared_conv/activate/Relu_output_0, 5_/task_heads.0/rot/rot.0/conv/Conv_output_0
 90,            Relu, 1, 1, 5_/task_heads.0/rot/rot.0/conv/Conv_output_0, 5_/task_heads.0/rot/rot.0/activate/Relu_output_0
 91,            Conv, 3, 1, 5_/task_heads.0/rot/rot.0/activate/Relu_output_0, 5_265
 92,            Conv, 3, 1, 5_/shared_conv/activate/Relu_output_0, 5_/task_heads.0/vel/vel.0/conv/Conv_output_0
 93,            Relu, 1, 1, 5_/task_heads.0/vel/vel.0/conv/Conv_output_0, 5_/task_heads.0/vel/vel.0/activate/Relu_output_0
 94,            Conv, 3, 1, 5_/task_heads.0/vel/vel.0/activate/Relu_output_0, 5_269
 95,            Conv, 3, 1, 5_/shared_conv/activate/Relu_output_0, 5_/task_heads.0/heatmap/heatmap.0/conv/Conv_output_0
 96,            Relu, 1, 1, 5_/task_heads.0/heatmap/heatmap.0/conv/Conv_output_0, 5_/task_heads.0/heatmap/heatmap.0/activate/Relu_output_0
 97,            Conv, 3, 1, 5_/task_heads.0/heatmap/heatmap.0/activate/Relu_output_0, 5_273

Input tensor name -  2_/Mul_output_0 

Input tensor name -  4_circle_map 

Input tensor name -  4_ray_map 

Input tensor name -  1_context 
Output tensor name - 5_273 
Output tensor name - 5_269 
Output tensor name - 5_265 
Output tensor name - 5_261 
Output tensor name - 5_257 
Output tensor name - 5_preds 
 Graph Domain TO version : 11In TIDL_onnxRtImportInit subgraph_name=subgraph_0
Layer 0, subgraph id subgraph_0, name=1_/Reshape_1_output_0
Layer 1, subgraph id subgraph_0, name=1_/context_se/gate/Sigmoid_output_0
Layer 2, subgraph id subgraph_0, name=1_/reduce_conv/reduce_conv.2/Relu_output_0
Layer 3, subgraph id subgraph_0, name=1_mlp_input
Layer 4, subgraph id subgraph_0, name=0_img
In TIDL_runtimesOptimizeNet: LayerIndex = 58, dataIndex = 55 
WARNING: Conv Layer 0_/0/layer2/layer2.0/conv1/Conv's coeff cannot be found(or not match) in coef file, Random coeff will be generated! Only for evaluation usage! Results are all random!
WARNING: Conv Layer 0_/0/layer2/layer2.0/downsample/downsample.0/Conv's coeff cannot be found(or not match) in coef file, Random coeff will be generated! Only for evaluation usage! Results are all random!
^CTraceback (most recent call last):
  File "/home/ht/edgeai/edgeai-tidl-tools/examples/osrt_python/ort/multi_input_no_post.py", line 331, in <module>
    nthreads = join_one(nthreads)
  File "/home/ht/edgeai/edgeai-tidl-tools/examples/osrt_python/ort/multi_input_no_post.py", line 313, in join_one
    sem.acquire()
KeyboardInterrupt
^CException ignored in atexit callback: <function _exit_function at 0x7a31795e9ea0>
Traceback (most recent call last):
  File "/home/ht/.pyenv/versions/3.10.14/lib/python3.10/multiprocessing/util.py", line 357, in _exit_function
    p.join()
  File "/home/ht/.pyenv/versions/3.10.14/lib/python3.10/multiprocessing/process.py", line 149, in join
    res = self._popen.wait(timeout)
  File "/home/ht/.pyenv/versions/3.10.14/lib/python3.10/multiprocessing/popen_fork.py", line 43, in wait
    return self.poll(os.WNOHANG if timeout == 0.0 else 0)
  File "/home/ht/.pyenv/versions/3.10.14/lib/python3.10/multiprocessing/popen_fork.py", line 27, in poll
    pid, sts = os.waitpid(self.pid, flag)
KeyboardInterrupt: 

(tidltools_09_02) ht@ht-OMEN:~/edgeai/edgeai-tidl-tools/examples/osrt_python/ort$

Do you have any clues or suggestion?

Regards,

Adam

0 Pratik Kedar over 1 year ago in reply to Adam Hua

TI__Mastermind 24041 points

Thanks for observations adam, from the logs seems like its stuck due to thread wait.

Can you set uncomment below line and try it again ?

https://github.com/TexasInstruments/edgeai-tidl-tools/blob/09_02_07_00/examples/osrt_python/ort/onnxrt_ep.py#L63

0 Adam Hua over 1 year ago in reply to Pratik Kedar

TI__Expert 5015 points

Hi Pratik,

cpu number has been already set to 1 since I only run one model. The process still stuck and the log is exactly the same.

0 Pratik Kedar over 1 year ago in reply to Adam Hua

TI__Mastermind 24041 points

Noted, let me connect with you on internal communication channel.

0 Adam Hua over 1 year ago in reply to Pratik Kedar

TI__Expert 5015 points

Hi Pratik,

I tried 32bit and it stuck at the same place, the log is exactly the same. This models runs successfully on onnx rt.

0 Pratik Kedar over 1 year ago in reply to Adam Hua

TI__Mastermind 24041 points

Noted

0 Adam Hua over 1 year ago in reply to Wu han

TI__Expert 5015 points

Hi han,

Please change this node. The mul layer is not supported. Mul should either be the same dimension for two inputs or should be 1x1 in HxW for one input and other dimension values the same for two inputs.

Regards,

Adam

0 Adam Hua over 1 year ago in reply to Wu han

TI__Expert 5015 points

Hi Han,

And please add this advanced options of c7x firmware version to enable broadcast Mul:

Regards,

Adam

0 Wu han over 1 year ago in reply to Adam Hua

Prodigy 10 points

I have added the options, but the state is same with before. and this is the model of same shape between 1_depth and 2_/Transpose_1_output_0.MatrixVT_model_2024-06-11-14-43-42_tda4.rar

0 Adam Hua over 1 year ago in reply to Wu han

TI__Expert 5015 points

Hi Han,

Thanks for your quick input. The process ends with "ERROR: Batch processing is not supported for this TIDL_ReduceMeanLayer, so exiting from import " on my side. This log comes from that the input tensors of the model indicates that batch processing is used. AM68a does not support batch processing at all so the batch dimension of all inputs should be 1.

Regards,

Adam

0 Wu han over 1 year ago in reply to Adam Hua

Prodigy 10 points

this is the model of input with 1, 18,256,704 tensor shape.MatrixVT_model_2024-06-12-11-40-51_tda4.rar

0 Adam Hua over 1 year ago in reply to Wu han

TI__Expert 5015 points

Hi Han,

The model can pass the first few operations in import now but still got some error. We will look into this further.

python3 multi_input_no_post.py -c -m matrix
Available execution providers :  ['TIDLExecutionProvider', 'TIDLCompilationProvider', 'CPUExecutionProvider']

Running 1 Models - ['motorvis_matrix']


Running_Model :  motorvis_matrix  


Running shape inference on model ../../../models/public/MatrixVT_model_2024-06-12-11-40-51_tda4.onnx 

tidl_tools_path                                 = /home/ht/edgeai/edgeai-tidl-tools/tidl_tools_am68a 
artifacts_folder                                = ../../../model-artifacts//motorvis_matrix/ 
tidl_tensor_bits                                = 8 
debug_level                                     = 6 
num_tidl_subgraphs                              = 16 
tidl_denylist                                   = 
tidl_denylist_layer_name                        = 
tidl_denylist_layer_type                         = 
tidl_allowlist_layer_name                        = 
model_type                                      =  
tidl_calibration_accuracy_level                 = 7 
tidl_calibration_options:num_frames_calibration = 2 
tidl_calibration_options:bias_calibration_iterations = 5 
mixed_precision_factor = -1.000000 
model_group_id = 0 
power_of_2_quantization                         = 2 
ONNX QDQ Enabled                                = 0 
enable_high_resolution_optimization             = 0 
pre_batchnorm_fold                              = 1 
add_data_convert_ops                            = 3 
output_feature_16bit_names_list                 =  
m_params_16bit_names_list                       =  
m_single_core_layers_names_list                    =  
reserved_compile_constraints_flag               = 1601 
ti_internal_reserved_1                          = 


 ****** WARNING : Network not identified as Object Detection network : (1) Ignore if network is not Object Detection network (2) If network is Object Detection network, please specify "model_type":"OD" as part of OSRT compilation options******

Supported TIDL layer type ---         Reshape -- 1_/Reshape 
Supported TIDL layer type ---            Gemm -- 1_/context_mlp/fc1/Gemm 
Supported TIDL layer type ---            Relu -- 1_/context_mlp/act/Relu 
Supported TIDL layer type ---            Gemm -- 1_/context_mlp/fc2/Gemm 
Supported TIDL layer type ---         Reshape -- 1_/Reshape_1 
Supported TIDL layer type ---            Conv -- 1_/context_se/conv_reduce/Conv 
Supported TIDL layer type ---            Relu -- 1_/context_se/act1/Relu 
Supported TIDL layer type ---            Conv -- 1_/context_se/conv_expand/Conv 
Supported TIDL layer type ---         Sigmoid -- 1_/context_se/gate/Sigmoid 
Supported TIDL layer type ---         Reshape -- 0_Reshape_Node 
Supported TIDL layer type ---            Conv -- 0_/0/conv1/Conv 
Supported TIDL layer type ---            Relu -- 0_/0/relu/Relu 
Supported TIDL layer type ---         MaxPool -- 0_/0/maxpool/MaxPool 
Supported TIDL layer type ---            Conv -- 0_/0/layer1/layer1.0/conv1/Conv 
Supported TIDL layer type ---            Relu -- 0_/0/layer1/layer1.0/relu/Relu 
Supported TIDL layer type ---            Conv -- 0_/0/layer1/layer1.0/conv2/Conv 
Supported TIDL layer type ---             Add -- 0_/0/layer1/layer1.0/Add 
Supported TIDL layer type ---            Relu -- 0_/0/layer1/layer1.0/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 0_/0/layer1/layer1.1/conv1/Conv 
Supported TIDL layer type ---            Relu -- 0_/0/layer1/layer1.1/relu/Relu 
Supported TIDL layer type ---            Conv -- 0_/0/layer1/layer1.1/conv2/Conv 
Supported TIDL layer type ---             Add -- 0_/0/layer1/layer1.1/Add 
Supported TIDL layer type ---            Relu -- 0_/0/layer1/layer1.1/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 0_/0/layer2/layer2.0/conv1/Conv 
Supported TIDL layer type ---            Relu -- 0_/0/layer2/layer2.0/relu/Relu 
Supported TIDL layer type ---            Conv -- 0_/0/layer2/layer2.0/conv2/Conv 
Supported TIDL layer type ---            Conv -- 0_/0/layer2/layer2.0/downsample/downsample.0/Conv 
Supported TIDL layer type ---             Add -- 0_/0/layer2/layer2.0/Add 
Supported TIDL layer type ---            Relu -- 0_/0/layer2/layer2.0/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 0_/0/layer2/layer2.1/conv1/Conv 
Supported TIDL layer type ---            Relu -- 0_/0/layer2/layer2.1/relu/Relu 
Supported TIDL layer type ---            Conv -- 0_/0/layer2/layer2.1/conv2/Conv 
Supported TIDL layer type ---             Add -- 0_/0/layer2/layer2.1/Add 
Supported TIDL layer type ---            Relu -- 0_/0/layer2/layer2.1/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 0_/0/layer3/layer3.0/conv1/Conv 
Supported TIDL layer type ---            Relu -- 0_/0/layer3/layer3.0/relu/Relu 
Supported TIDL layer type ---            Conv -- 0_/0/layer3/layer3.0/conv2/Conv 
Supported TIDL layer type ---            Conv -- 0_/0/layer3/layer3.0/downsample/downsample.0/Conv 
Supported TIDL layer type ---             Add -- 0_/0/layer3/layer3.0/Add 
Supported TIDL layer type ---            Relu -- 0_/0/layer3/layer3.0/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 0_/0/layer3/layer3.1/conv1/Conv 
Supported TIDL layer type ---            Relu -- 0_/0/layer3/layer3.1/relu/Relu 
Supported TIDL layer type ---            Conv -- 0_/0/layer3/layer3.1/conv2/Conv 
Supported TIDL layer type ---             Add -- 0_/0/layer3/layer3.1/Add 
Supported TIDL layer type ---            Relu -- 0_/0/layer3/layer3.1/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 0_/1/deblocks.2/deblocks.2.0/Conv 
Supported TIDL layer type ---            Relu -- 0_/1/deblocks.2/deblocks.2.2/Relu 
Supported TIDL layer type ---            Conv -- 0_/1/deblocks.1/deblocks.1.0/Conv 
Supported TIDL layer type ---            Relu -- 0_/1/deblocks.1/deblocks.1.2/Relu 
Supported TIDL layer type ---            Conv -- 0_/1/deblocks.0/deblocks.0.0/Conv 
Supported TIDL layer type ---            Relu -- 0_/1/deblocks.0/deblocks.0.2/Relu 
Supported TIDL layer type ---          Concat -- 0_/1/Concat 
Supported TIDL layer type ---            Conv -- 1_/reduce_conv/reduce_conv.0/Conv 
Supported TIDL layer type ---            Relu -- 1_/reduce_conv/reduce_conv.2/Relu 
Supported TIDL layer type ---             Mul -- 1_/context_se/Mul 
Supported TIDL layer type ---            Conv -- 1_/context_conv/Conv 
Supported TIDL layer type ---            Conv -- 2_/vertical_weighter/vertical_weighter.0/Conv 
Supported TIDL layer type ---            Relu -- 2_/vertical_weighter/vertical_weighter.2/Relu 
Supported TIDL layer type ---            Conv -- 2_/vertical_weighter/vertical_weighter.3/Conv 
Supported TIDL layer type ---       Transpose -- 2_/Transpose 
Supported TIDL layer type ---         Softmax -- 2_/Softmax 
Supported TIDL layer type ---       Transpose -- 2_/Transpose_1 
Supported TIDL layer type ---            Conv -- 1_/depth_se/conv_reduce/Conv 
Supported TIDL layer type ---            Relu -- 1_/depth_se/act1/Relu 
Supported TIDL layer type ---            Conv -- 1_/depth_se/conv_expand/Conv 
Supported TIDL layer type ---         Sigmoid -- 1_/depth_se/gate/Sigmoid 
Supported TIDL layer type ---             Mul -- 1_/depth_se/Mul 
Supported TIDL layer type ---            Conv -- 1_/depth_conv/depth_conv.0/conv1/Conv 
Supported TIDL layer type ---            Relu -- 1_/depth_conv/depth_conv.0/relu/Relu 
Supported TIDL layer type ---            Conv -- 1_/depth_conv/depth_conv.0/conv2/Conv 
Supported TIDL layer type ---             Add -- 1_/depth_conv/depth_conv.0/Add 
Supported TIDL layer type ---            Relu -- 1_/depth_conv/depth_conv.0/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 1_/depth_conv/depth_conv.1/conv1/Conv 
Supported TIDL layer type ---            Relu -- 1_/depth_conv/depth_conv.1/relu/Relu 
Supported TIDL layer type ---            Conv -- 1_/depth_conv/depth_conv.1/conv2/Conv 
Supported TIDL layer type ---             Add -- 1_/depth_conv/depth_conv.1/Add 
Supported TIDL layer type ---            Relu -- 1_/depth_conv/depth_conv.1/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 1_/depth_conv/depth_conv.2/conv1/Conv 
Supported TIDL layer type ---            Relu -- 1_/depth_conv/depth_conv.2/relu/Relu 
Supported TIDL layer type ---            Conv -- 1_/depth_conv/depth_conv.2/conv2/Conv 
Supported TIDL layer type ---             Add -- 1_/depth_conv/depth_conv.2/Add 
Supported TIDL layer type ---            Relu -- 1_/depth_conv/depth_conv.2/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 1_/depth_conv/depth_conv.3/Conv 
Supported TIDL layer type ---         Softmax -- 1_/Softmax 
Supported TIDL layer type ---             Mul -- 2_/Mul 
Supported TIDL layer type ---       Transpose -- 2_/Transpose_2 
Supported TIDL layer type ---         Reshape -- 2_/Reshape 
Supported TIDL layer type ---      ReduceMean -- 2_/ReduceMean 
Supported TIDL layer type ---          MatMul -- 4_/MatMul 
Supported TIDL layer type ---             Mul -- 4_/Mul 
Supported TIDL layer type ---       Transpose -- 4_/Transpose_1 
Supported TIDL layer type ---            Conv -- 3_/context_conv/Conv 
Supported TIDL layer type ---            Conv -- 3_/merger/merger.0/Conv 
Supported TIDL layer type ---         Sigmoid -- 3_/merger/merger.1/Sigmoid 
Supported TIDL layer type ---            Conv -- 3_/merger/merger.2/Conv 
Supported TIDL layer type ---            Conv -- 3_/reduce_conv/reduce_conv.0/Conv 
Supported TIDL layer type ---            Relu -- 3_/reduce_conv/reduce_conv.2/Relu 
Supported TIDL layer type ---            Conv -- 3_/conv1/conv1.0/Conv 
Supported TIDL layer type ---            Relu -- 3_/conv1/conv1.2/Relu 
Supported TIDL layer type ---            Conv -- 3_/conv1/conv1.3/Conv 
Supported TIDL layer type ---            Relu -- 3_/conv1/conv1.5/Relu 
Supported TIDL layer type ---             Add -- 3_/Add 
Supported TIDL layer type ---            Conv -- 3_/conv2/conv2.0/Conv 
Supported TIDL layer type ---            Relu -- 3_/conv2/conv2.2/Relu 
Supported TIDL layer type ---            Conv -- 3_/conv2/conv2.3/Conv 
Supported TIDL layer type ---            Relu -- 3_/conv2/conv2.5/Relu 
Supported TIDL layer type ---             Add -- 3_/Add_1 
Supported TIDL layer type ---            Conv -- 3_/out_conv/out_conv.0/Conv 
Supported TIDL layer type ---            Relu -- 3_/out_conv/out_conv.2/Relu 
Supported TIDL layer type ---       Transpose -- 3_/Transpose 
Supported TIDL layer type ---            Conv -- 3_/gather_conv/Conv 
Supported TIDL layer type ---         Reshape -- 3_/Reshape 
Supported TIDL layer type ---       Transpose -- 4_/Transpose 
Supported TIDL layer type ---         Reshape -- 4_/Reshape 
Supported TIDL layer type ---          MatMul -- 4_/MatMul_1 
Supported TIDL layer type ---       Transpose -- 4_/Transpose_2 
Supported TIDL layer type ---         Reshape -- 4_/Reshape_1 
Supported TIDL layer type ---            Conv -- 5_/conv1/Conv 
Supported TIDL layer type ---            Relu -- 5_/relu/Relu 
Supported TIDL layer type ---            Conv -- 5_/layer1/layer1.0/conv1/Conv 
Supported TIDL layer type ---            Relu -- 5_/layer1/layer1.0/relu/Relu 
Supported TIDL layer type ---            Conv -- 5_/layer1/layer1.0/conv2/Conv 
Supported TIDL layer type ---             Add -- 5_/layer1/layer1.0/Add 
Supported TIDL layer type ---            Relu -- 5_/layer1/layer1.0/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 5_/layer1/layer1.1/conv1/Conv 
Supported TIDL layer type ---            Relu -- 5_/layer1/layer1.1/relu/Relu 
Supported TIDL layer type ---            Conv -- 5_/layer1/layer1.1/conv2/Conv 
Supported TIDL layer type ---             Add -- 5_/layer1/layer1.1/Add 
Supported TIDL layer type ---            Relu -- 5_/layer1/layer1.1/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 5_/layer2/layer2.0/downsample/downsample.0/Conv 
Supported TIDL layer type ---            Conv -- 5_/layer2/layer2.0/conv1/Conv 
Supported TIDL layer type ---            Relu -- 5_/layer2/layer2.0/relu/Relu 
Supported TIDL layer type ---            Conv -- 5_/layer2/layer2.0/conv2/Conv 
Supported TIDL layer type ---             Add -- 5_/layer2/layer2.0/Add 
Supported TIDL layer type ---            Relu -- 5_/layer2/layer2.0/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 5_/layer2/layer2.1/conv1/Conv 
Supported TIDL layer type ---            Relu -- 5_/layer2/layer2.1/relu/Relu 
Supported TIDL layer type ---            Conv -- 5_/layer2/layer2.1/conv2/Conv 
Supported TIDL layer type ---             Add -- 5_/layer2/layer2.1/Add 
Supported TIDL layer type ---            Relu -- 5_/layer2/layer2.1/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 5_/layer3/layer3.0/downsample/downsample.0/Conv 
Supported TIDL layer type ---            Conv -- 5_/layer3/layer3.0/conv1/Conv 
Supported TIDL layer type ---            Relu -- 5_/layer3/layer3.0/relu/Relu 
Supported TIDL layer type ---            Conv -- 5_/layer3/layer3.0/conv2/Conv 
Supported TIDL layer type ---             Add -- 5_/layer3/layer3.0/Add 
Supported TIDL layer type ---            Relu -- 5_/layer3/layer3.0/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 5_/layer3/layer3.1/conv1/Conv 
Supported TIDL layer type ---            Relu -- 5_/layer3/layer3.1/relu/Relu 
Supported TIDL layer type ---            Conv -- 5_/layer3/layer3.1/conv2/Conv 
Supported TIDL layer type ---             Add -- 5_/layer3/layer3.1/Add 
Supported TIDL layer type ---            Relu -- 5_/layer3/layer3.1/relu_1/Relu 
Supported TIDL layer type ---            Conv -- 5_/neck/deblocks.3/deblocks.3.0/Conv 
Supported TIDL layer type ---            Relu -- 5_/neck/deblocks.3/deblocks.3.2/Relu 
Supported TIDL layer type ---          Resize -- 5_/neck/Resize_2 
Supported TIDL layer type ---            Conv -- 5_/neck/deblocks.2/deblocks.2.0/Conv 
Supported TIDL layer type ---            Relu -- 5_/neck/deblocks.2/deblocks.2.2/Relu 
Supported TIDL layer type ---          Resize -- 5_/neck/Resize_1 
Supported TIDL layer type ---            Conv -- 5_/neck/deblocks.1/deblocks.1.0/Conv 
Supported TIDL layer type ---            Relu -- 5_/neck/deblocks.1/deblocks.1.2/Relu 
Supported TIDL layer type ---          Resize -- 5_/neck/Resize 
Supported TIDL layer type ---            Conv -- 5_/neck/deblocks.0/deblocks.0.0/Conv 
Supported TIDL layer type ---            Relu -- 5_/neck/deblocks.0/deblocks.0.2/Relu 
Supported TIDL layer type ---          Concat -- 5_/neck/Concat 
Supported TIDL layer type ---            Conv -- 5_/shared_conv/conv/Conv 
Supported TIDL layer type ---            Relu -- 5_/shared_conv/activate/Relu 
Supported TIDL layer type ---            Conv -- 5_/task_heads.0/heatmap/heatmap.0/conv/Conv 
Supported TIDL layer type ---            Relu -- 5_/task_heads.0/heatmap/heatmap.0/activate/Relu 
Supported TIDL layer type ---            Conv -- 5_/task_heads.0/heatmap/heatmap.1/Conv 
Supported TIDL layer type ---            Conv -- 5_/task_heads.0/vel/vel.0/conv/Conv 
Supported TIDL layer type ---            Relu -- 5_/task_heads.0/vel/vel.0/activate/Relu 
Supported TIDL layer type ---            Conv -- 5_/task_heads.0/vel/vel.1/Conv 
Supported TIDL layer type ---            Conv -- 5_/task_heads.0/rot/rot.0/conv/Conv 
Supported TIDL layer type ---            Relu -- 5_/task_heads.0/rot/rot.0/activate/Relu 
Supported TIDL layer type ---            Conv -- 5_/task_heads.0/rot/rot.1/Conv 
Supported TIDL layer type ---            Conv -- 5_/task_heads.0/dim/dim.0/conv/Conv 
Supported TIDL layer type ---            Relu -- 5_/task_heads.0/dim/dim.0/activate/Relu 
Supported TIDL layer type ---            Conv -- 5_/task_heads.0/dim/dim.1/Conv 
Supported TIDL layer type ---            Conv -- 5_/task_heads.0/height/height.0/conv/Conv 
Supported TIDL layer type ---            Relu -- 5_/task_heads.0/height/height.0/activate/Relu 
Supported TIDL layer type ---            Conv -- 5_/task_heads.0/height/height.1/Conv 
Supported TIDL layer type ---            Conv -- 5_/task_heads.0/reg/reg.0/conv/Conv 
Supported TIDL layer type ---            Relu -- 5_/task_heads.0/reg/reg.0/activate/Relu 
Supported TIDL layer type ---            Conv -- 5_/task_heads.0/reg/reg.1/Conv 

Preliminary subgraphs created = 1 
Final number of subgraphs created are : 1, - Offloaded Nodes - 183, Total Nodes - 183 
SUGGESTION -- [TIDL_ResizeLayer]  Resize kernel with non-power of 2 resize ratio is not optimal.  
INFORMATION -- [TIDL_ResizeLayer]  Any resize ratio which is power of 2 and greater than 4 will be placed by combination of 4x4 resize layer and 2x2 resize layer. For example a 8x8 resize will be replaced by 4x4 resize followed by 2x2 resize.  
INFORMATION -- [TIDL_ResizeLayer]  Any resize ratio which is power of 2 and greater than 4 will be placed by combination of 4x4 resize layer and 2x2 resize layer. For example a 8x8 resize will be replaced by 4x4 resize followed by 2x2 resize.  
Running runtimes graphviz - /home/ht/edgeai/edgeai-tidl-tools/tidl_tools_am68a/tidl_graphVisualiser_runtimes.out ../../../model-artifacts//matrix//allowedNode.txt ../../../model-artifacts//motorvis_matrix//tempDir/graphvizInfo.txt ../../../model-artifacts//motorvis_matrix//tempDir/runtimes_visualization.svg 
*** In TIDL_createStateImportFunc *** 
Compute on node : TIDLExecutionProvider_TIDL_0_0
  0,         Reshape, 2, 1, 0_img, 0_reshaped_output
  1,            Conv, 3, 1, 0_reshaped_output, 0_/0/conv1/Conv_output_0
  2,            Relu, 1, 1, 0_/0/conv1/Conv_output_0, 0_/0/relu/Relu_output_0
  3,         MaxPool, 1, 1, 0_/0/relu/Relu_output_0, 0_/0/maxpool/MaxPool_output_0
  4,            Conv, 3, 1, 0_/0/maxpool/MaxPool_output_0, 0_/0/layer1/layer1.0/conv1/Conv_output_0
  5,            Relu, 1, 1, 0_/0/layer1/layer1.0/conv1/Conv_output_0, 0_/0/layer1/layer1.0/relu/Relu_output_0
  6,            Conv, 3, 1, 0_/0/layer1/layer1.0/relu/Relu_output_0, 0_/0/layer1/layer1.0/conv2/Conv_output_0
  7,             Add, 2, 1, 0_/0/layer1/layer1.0/conv2/Conv_output_0, 0_/0/layer1/layer1.0/Add_output_0
  8,            Relu, 1, 1, 0_/0/layer1/layer1.0/Add_output_0, 0_/0/layer1/layer1.0/relu_1/Relu_output_0
  9,            Conv, 3, 1, 0_/0/layer1/layer1.0/relu_1/Relu_output_0, 0_/0/layer1/layer1.1/conv1/Conv_output_0
 10,            Relu, 1, 1, 0_/0/layer1/layer1.1/conv1/Conv_output_0, 0_/0/layer1/layer1.1/relu/Relu_output_0
 11,            Conv, 3, 1, 0_/0/layer1/layer1.1/relu/Relu_output_0, 0_/0/layer1/layer1.1/conv2/Conv_output_0
 12,             Add, 2, 1, 0_/0/layer1/layer1.1/conv2/Conv_output_0, 0_/0/layer1/layer1.1/Add_output_0
 13,            Relu, 1, 1, 0_/0/layer1/layer1.1/Add_output_0, 0_/0/layer1/layer1.1/relu_1/Relu_output_0
 14,            Conv, 3, 1, 0_/0/layer1/layer1.1/relu_1/Relu_output_0, 0_/1/deblocks.0/deblocks.0.0/Conv_output_0
 15,            Relu, 1, 1, 0_/1/deblocks.0/deblocks.0.0/Conv_output_0, 0_/1/deblocks.0/deblocks.0.2/Relu_output_0
 16,            Conv, 3, 1, 0_/0/layer1/layer1.1/relu_1/Relu_output_0, 0_/0/layer2/layer2.0/downsample/downsample.0/Conv_output_0
 17,            Conv, 3, 1, 0_/0/layer1/layer1.1/relu_1/Relu_output_0, 0_/0/layer2/layer2.0/conv1/Conv_output_0
 18,            Relu, 1, 1, 0_/0/layer2/layer2.0/conv1/Conv_output_0, 0_/0/layer2/layer2.0/relu/Relu_output_0
 19,            Conv, 3, 1, 0_/0/layer2/layer2.0/relu/Relu_output_0, 0_/0/layer2/layer2.0/conv2/Conv_output_0
 20,             Add, 2, 1, 0_/0/layer2/layer2.0/conv2/Conv_output_0, 0_/0/layer2/layer2.0/Add_output_0
 21,            Relu, 1, 1, 0_/0/layer2/layer2.0/Add_output_0, 0_/0/layer2/layer2.0/relu_1/Relu_output_0
 22,            Conv, 3, 1, 0_/0/layer2/layer2.0/relu_1/Relu_output_0, 0_/0/layer2/layer2.1/conv1/Conv_output_0
 23,            Relu, 1, 1, 0_/0/layer2/layer2.1/conv1/Conv_output_0, 0_/0/layer2/layer2.1/relu/Relu_output_0
 24,            Conv, 3, 1, 0_/0/layer2/layer2.1/relu/Relu_output_0, 0_/0/layer2/layer2.1/conv2/Conv_output_0
 25,             Add, 2, 1, 0_/0/layer2/layer2.1/conv2/Conv_output_0, 0_/0/layer2/layer2.1/Add_output_0
 26,            Relu, 1, 1, 0_/0/layer2/layer2.1/Add_output_0, 0_/0/layer2/layer2.1/relu_1/Relu_output_0
 27,            Conv, 3, 1, 0_/0/layer2/layer2.1/relu_1/Relu_output_0, 0_/1/deblocks.1/deblocks.1.0/Conv_output_0
 28,            Relu, 1, 1, 0_/1/deblocks.1/deblocks.1.0/Conv_output_0, 0_/1/deblocks.1/deblocks.1.2/Relu_output_0
 29,            Conv, 3, 1, 0_/0/layer2/layer2.1/relu_1/Relu_output_0, 0_/0/layer3/layer3.0/downsample/downsample.0/Conv_output_0
 30,            Conv, 3, 1, 0_/0/layer2/layer2.1/relu_1/Relu_output_0, 0_/0/layer3/layer3.0/conv1/Conv_output_0
 31,            Relu, 1, 1, 0_/0/layer3/layer3.0/conv1/Conv_output_0, 0_/0/layer3/layer3.0/relu/Relu_output_0
 32,            Conv, 3, 1, 0_/0/layer3/layer3.0/relu/Relu_output_0, 0_/0/layer3/layer3.0/conv2/Conv_output_0
 33,             Add, 2, 1, 0_/0/layer3/layer3.0/conv2/Conv_output_0, 0_/0/layer3/layer3.0/Add_output_0
 34,            Relu, 1, 1, 0_/0/layer3/layer3.0/Add_output_0, 0_/0/layer3/layer3.0/relu_1/Relu_output_0
 35,            Conv, 3, 1, 0_/0/layer3/layer3.0/relu_1/Relu_output_0, 0_/0/layer3/layer3.1/conv1/Conv_output_0
 36,            Relu, 1, 1, 0_/0/layer3/layer3.1/conv1/Conv_output_0, 0_/0/layer3/layer3.1/relu/Relu_output_0
 37,            Conv, 3, 1, 0_/0/layer3/layer3.1/relu/Relu_output_0, 0_/0/layer3/layer3.1/conv2/Conv_output_0
 38,             Add, 2, 1, 0_/0/layer3/layer3.1/conv2/Conv_output_0, 0_/0/layer3/layer3.1/Add_output_0
 39,            Relu, 1, 1, 0_/0/layer3/layer3.1/Add_output_0, 0_/0/layer3/layer3.1/relu_1/Relu_output_0
 40,            Conv, 3, 1, 0_/0/layer3/layer3.1/relu_1/Relu_output_0, 0_/1/deblocks.2/deblocks.2.0/Conv_output_0
 41,            Relu, 1, 1, 0_/1/deblocks.2/deblocks.2.0/Conv_output_0, 0_/1/deblocks.2/deblocks.2.2/Relu_output_0
 42,          Concat, 3, 1, 0_/1/deblocks.0/deblocks.0.2/Relu_output_0, 0_backbone_feature
 43,            Conv, 3, 1, 0_backbone_feature, 1_/reduce_conv/reduce_conv.0/Conv_output_0
 44,            Relu, 1, 1, 1_/reduce_conv/reduce_conv.0/Conv_output_0, 1_/reduce_conv/reduce_conv.2/Relu_output_0
 45,         Reshape, 2, 1, 1_mlp_input, 1_/Reshape_output_0
 46,            Gemm, 3, 1, 1_/Reshape_output_0, 1_/context_mlp/fc1/Gemm_output_0
 47,            Relu, 1, 1, 1_/context_mlp/fc1/Gemm_output_0, 1_/context_mlp/act/Relu_output_0
 48,            Gemm, 3, 1, 1_/context_mlp/act/Relu_output_0, 1_/context_mlp/fc2/Gemm_output_0
 49,         Reshape, 2, 1, 1_/context_mlp/fc2/Gemm_output_0, 1_/Reshape_1_output_0
 50,            Conv, 3, 1, 1_/Reshape_1_output_0, 1_/context_se/conv_reduce/Conv_output_0
 51,            Relu, 1, 1, 1_/context_se/conv_reduce/Conv_output_0, 1_/context_se/act1/Relu_output_0
 52,            Conv, 3, 1, 1_/context_se/act1/Relu_output_0, 1_/context_se/conv_expand/Conv_output_0
 53,         Sigmoid, 1, 1, 1_/context_se/conv_expand/Conv_output_0, 1_/context_se/gate/Sigmoid_output_0
 54,             Mul, 2, 1, 1_/reduce_conv/reduce_conv.2/Relu_output_0, 1_/context_se/Mul_output_0
 55,            Conv, 3, 1, 1_/context_se/Mul_output_0, 1_context
 56,            Conv, 3, 1, 1_context, 3_/context_conv/Conv_output_0
 57,            Conv, 3, 1, 3_/context_conv/Conv_output_0, 3_/merger/merger.0/Conv_output_0
 58,         Sigmoid, 1, 1, 3_/merger/merger.0/Conv_output_0, 3_/merger/merger.1/Sigmoid_output_0
 59,            Conv, 3, 1, 3_/merger/merger.1/Sigmoid_output_0, 3_/merger/merger.2/Conv_output_0
 60,            Conv, 3, 1, 3_/merger/merger.2/Conv_output_0, 3_/reduce_conv/reduce_conv.0/Conv_output_0
 61,            Relu, 1, 1, 3_/reduce_conv/reduce_conv.0/Conv_output_0, 3_/reduce_conv/reduce_conv.2/Relu_output_0
 62,            Conv, 3, 1, 3_/reduce_conv/reduce_conv.2/Relu_output_0, 3_/conv1/conv1.0/Conv_output_0
 63,            Relu, 1, 1, 3_/conv1/conv1.0/Conv_output_0, 3_/conv1/conv1.2/Relu_output_0
 64,            Conv, 3, 1, 3_/conv1/conv1.2/Relu_output_0, 3_/conv1/conv1.3/Conv_output_0
 65,            Relu, 1, 1, 3_/conv1/conv1.3/Conv_output_0, 3_/conv1/conv1.5/Relu_output_0
 66,             Add, 2, 1, 3_/conv1/conv1.5/Relu_output_0, 3_/Add_output_0
 67,            Conv, 3, 1, 3_/Add_output_0, 3_/conv2/conv2.0/Conv_output_0
 68,            Relu, 1, 1, 3_/conv2/conv2.0/Conv_output_0, 3_/conv2/conv2.2/Relu_output_0
 69,            Conv, 3, 1, 3_/conv2/conv2.2/Relu_output_0, 3_/conv2/conv2.3/Conv_output_0
 70,            Relu, 1, 1, 3_/conv2/conv2.3/Conv_output_0, 3_/conv2/conv2.5/Relu_output_0
 71,             Add, 2, 1, 3_/conv2/conv2.5/Relu_output_0, 3_/Add_1_output_0
 72,            Conv, 3, 1, 3_/Add_1_output_0, 3_/out_conv/out_conv.0/Conv_output_0
 73,            Relu, 1, 1, 3_/out_conv/out_conv.0/Conv_output_0, 3_/out_conv/out_conv.2/Relu_output_0
 74,       Transpose, 1, 1, 3_/out_conv/out_conv.2/Relu_output_0, 3_/Transpose_output_0
 75,            Conv, 3, 1, 3_/Transpose_output_0, 3_/gather_conv/Conv_output_0
 76,         Reshape, 2, 1, 3_/gather_conv/Conv_output_0, 3_HoriConv_out
 77,       Transpose, 1, 1, 3_HoriConv_out, 4_/Transpose_output_0
 78,         Reshape, 2, 1, 4_/Transpose_output_0, 4_/Reshape_output_0
 79,            Conv, 3, 1, 1_/Reshape_1_output_0, 1_/depth_se/conv_reduce/Conv_output_0
 80,            Relu, 1, 1, 1_/depth_se/conv_reduce/Conv_output_0, 1_/depth_se/act1/Relu_output_0
 81,            Conv, 3, 1, 1_/depth_se/act1/Relu_output_0, 1_/depth_se/conv_expand/Conv_output_0
 82,         Sigmoid, 1, 1, 1_/depth_se/conv_expand/Conv_output_0, 1_/depth_se/gate/Sigmoid_output_0
 83,             Mul, 2, 1, 1_/reduce_conv/reduce_conv.2/Relu_output_0, 1_/depth_se/Mul_output_0
 84,            Conv, 3, 1, 1_/depth_se/Mul_output_0, 1_/depth_conv/depth_conv.0/conv1/Conv_output_0
 85,            Relu, 1, 1, 1_/depth_conv/depth_conv.0/conv1/Conv_output_0, 1_/depth_conv/depth_conv.0/relu/Relu_output_0
 86,            Conv, 3, 1, 1_/depth_conv/depth_conv.0/relu/Relu_output_0, 1_/depth_conv/depth_conv.0/conv2/Conv_output_0
 87,             Add, 2, 1, 1_/depth_conv/depth_conv.0/conv2/Conv_output_0, 1_/depth_conv/depth_conv.0/Add_output_0
 88,            Relu, 1, 1, 1_/depth_conv/depth_conv.0/Add_output_0, 1_/depth_conv/depth_conv.0/relu_1/Relu_output_0
 89,            Conv, 3, 1, 1_/depth_conv/depth_conv.0/relu_1/Relu_output_0, 1_/depth_conv/depth_conv.1/conv1/Conv_output_0
 90,            Relu, 1, 1, 1_/depth_conv/depth_conv.1/conv1/Conv_output_0, 1_/depth_conv/depth_conv.1/relu/Relu_output_0
 91,            Conv, 3, 1, 1_/depth_conv/depth_conv.1/relu/Relu_output_0, 1_/depth_conv/depth_conv.1/conv2/Conv_output_0
 92,             Add, 2, 1, 1_/depth_conv/depth_conv.1/conv2/Conv_output_0, 1_/depth_conv/depth_conv.1/Add_output_0
 93,            Relu, 1, 1, 1_/depth_conv/depth_conv.1/Add_output_0, 1_/depth_conv/depth_conv.1/relu_1/Relu_output_0
 94,            Conv, 3, 1, 1_/depth_conv/depth_conv.1/relu_1/Relu_output_0, 1_/depth_conv/depth_conv.2/conv1/Conv_output_0
 95,            Relu, 1, 1, 1_/depth_conv/depth_conv.2/conv1/Conv_output_0, 1_/depth_conv/depth_conv.2/relu/Relu_output_0
 96,            Conv, 3, 1, 1_/depth_conv/depth_conv.2/relu/Relu_output_0, 1_/depth_conv/depth_conv.2/conv2/Conv_output_0
 97,             Add, 2, 1, 1_/depth_conv/depth_conv.2/conv2/Conv_output_0, 1_/depth_conv/depth_conv.2/Add_output_0
 98,            Relu, 1, 1, 1_/depth_conv/depth_conv.2/Add_output_0, 1_/depth_conv/depth_conv.2/relu_1/Relu_output_0
 99,            Conv, 3, 1, 1_/depth_conv/depth_conv.2/relu_1/Relu_output_0, 1_/depth_conv/depth_conv.3/Conv_output_0
100,         Softmax, 1, 1, 1_/depth_conv/depth_conv.3/Conv_output_0, 1_depth
101,            Conv, 3, 1, 1_context, 2_/vertical_weighter/vertical_weighter.0/Conv_output_0
102,            Relu, 1, 1, 2_/vertical_weighter/vertical_weighter.0/Conv_output_0, 2_/vertical_weighter/vertical_weighter.2/Relu_output_0
103,            Conv, 3, 1, 2_/vertical_weighter/vertical_weighter.2/Relu_output_0, 2_/vertical_weighter/vertical_weighter.3/Conv_output_0
104,       Transpose, 1, 1, 2_/vertical_weighter/vertical_weighter.3/Conv_output_0, 2_/Transpose_output_0
105,         Softmax, 1, 1, 2_/Transpose_output_0, 2_/Softmax_output_0
106,       Transpose, 1, 1, 2_/Softmax_output_0, 2_/Transpose_1_output_0
107,             Mul, 2, 1, 1_depth, 2_/Mul_output_0
108,       Transpose, 1, 1, 2_/Mul_output_0, 2_/Transpose_2_output_0
109,         Reshape, 2, 1, 2_/Transpose_2_output_0, 2_/Reshape_output_0
110,      ReduceMean, 1, 1, 2_/Reshape_output_0, 2_DepthReducer_out
111,          MatMul, 2, 1, 2_DepthReducer_out, 4_/MatMul_output_0
112,             Mul, 2, 1, 4_/MatMul_output_0, 4_/Mul_output_0
113,       Transpose, 1, 1, 4_/Mul_output_0, 4_/Transpose_1_output_0
114,          MatMul, 2, 1, 4_/Transpose_1_output_0, 4_/MatMul_1_output_0
115,       Transpose, 1, 1, 4_/MatMul_1_output_0, 4_/Transpose_2_output_0
116,         Reshape, 2, 1, 4_/Transpose_2_output_0, 4_bev_out
117,            Conv, 3, 1, 4_bev_out, 5_/neck/deblocks.0/deblocks.0.0/Conv_output_0
118,            Relu, 1, 1, 5_/neck/deblocks.0/deblocks.0.0/Conv_output_0, 5_/neck/deblocks.0/deblocks.0.2/Relu_output_0
119,            Conv, 3, 1, 4_bev_out, 5_/conv1/Conv_output_0
120,            Relu, 1, 1, 5_/conv1/Conv_output_0, 5_/relu/Relu_output_0
121,            Conv, 3, 1, 5_/relu/Relu_output_0, 5_/layer1/layer1.0/conv1/Conv_output_0
122,            Relu, 1, 1, 5_/layer1/layer1.0/conv1/Conv_output_0, 5_/layer1/layer1.0/relu/Relu_output_0
123,            Conv, 3, 1, 5_/layer1/layer1.0/relu/Relu_output_0, 5_/layer1/layer1.0/conv2/Conv_output_0
124,             Add, 2, 1, 5_/layer1/layer1.0/conv2/Conv_output_0, 5_/layer1/layer1.0/Add_output_0
125,            Relu, 1, 1, 5_/layer1/layer1.0/Add_output_0, 5_/layer1/layer1.0/relu_1/Relu_output_0
126,            Conv, 3, 1, 5_/layer1/layer1.0/relu_1/Relu_output_0, 5_/layer1/layer1.1/conv1/Conv_output_0
127,            Relu, 1, 1, 5_/layer1/layer1.1/conv1/Conv_output_0, 5_/layer1/layer1.1/relu/Relu_output_0
128,            Conv, 3, 1, 5_/layer1/layer1.1/relu/Relu_output_0, 5_/layer1/layer1.1/conv2/Conv_output_0
129,             Add, 2, 1, 5_/layer1/layer1.1/conv2/Conv_output_0, 5_/layer1/layer1.1/Add_output_0
130,            Relu, 1, 1, 5_/layer1/layer1.1/Add_output_0, 5_/layer1/layer1.1/relu_1/Relu_output_0
131,            Conv, 3, 1, 5_/layer1/layer1.1/relu_1/Relu_output_0, 5_/neck/deblocks.1/deblocks.1.0/Conv_output_0
132,            Relu, 1, 1, 5_/neck/deblocks.1/deblocks.1.0/Conv_output_0, 5_/neck/deblocks.1/deblocks.1.2/Relu_output_0
133,          Resize, 3, 1, 5_/neck/deblocks.1/deblocks.1.2/Relu_output_0, 5_/neck/Resize_output_0
134,            Conv, 3, 1, 5_/layer1/layer1.1/relu_1/Relu_output_0, 5_/layer2/layer2.0/conv1/Conv_output_0
135,            Relu, 1, 1, 5_/layer2/layer2.0/conv1/Conv_output_0, 5_/layer2/layer2.0/relu/Relu_output_0
136,            Conv, 3, 1, 5_/layer2/layer2.0/relu/Relu_output_0, 5_/layer2/layer2.0/conv2/Conv_output_0
137,            Conv, 3, 1, 5_/layer1/layer1.1/relu_1/Relu_output_0, 5_/layer2/layer2.0/downsample/downsample.0/Conv_output_0
138,             Add, 2, 1, 5_/layer2/layer2.0/conv2/Conv_output_0, 5_/layer2/layer2.0/Add_output_0
139,            Relu, 1, 1, 5_/layer2/layer2.0/Add_output_0, 5_/layer2/layer2.0/relu_1/Relu_output_0
140,            Conv, 3, 1, 5_/layer2/layer2.0/relu_1/Relu_output_0, 5_/layer2/layer2.1/conv1/Conv_output_0
141,            Relu, 1, 1, 5_/layer2/layer2.1/conv1/Conv_output_0, 5_/layer2/layer2.1/relu/Relu_output_0
142,            Conv, 3, 1, 5_/layer2/layer2.1/relu/Relu_output_0, 5_/layer2/layer2.1/conv2/Conv_output_0
143,             Add, 2, 1, 5_/layer2/layer2.1/conv2/Conv_output_0, 5_/layer2/layer2.1/Add_output_0
144,            Relu, 1, 1, 5_/layer2/layer2.1/Add_output_0, 5_/layer2/layer2.1/relu_1/Relu_output_0
145,            Conv, 3, 1, 5_/layer2/layer2.1/relu_1/Relu_output_0, 5_/neck/deblocks.2/deblocks.2.0/Conv_output_0
146,            Relu, 1, 1, 5_/neck/deblocks.2/deblocks.2.0/Conv_output_0, 5_/neck/deblocks.2/deblocks.2.2/Relu_output_0
147,          Resize, 3, 1, 5_/neck/deblocks.2/deblocks.2.2/Relu_output_0, 5_/neck/Resize_1_output_0
148,            Conv, 3, 1, 5_/layer2/layer2.1/relu_1/Relu_output_0, 5_/layer3/layer3.0/conv1/Conv_output_0
149,            Relu, 1, 1, 5_/layer3/layer3.0/conv1/Conv_output_0, 5_/layer3/layer3.0/relu/Relu_output_0
150,            Conv, 3, 1, 5_/layer3/layer3.0/relu/Relu_output_0, 5_/layer3/layer3.0/conv2/Conv_output_0
151,            Conv, 3, 1, 5_/layer2/layer2.1/relu_1/Relu_output_0, 5_/layer3/layer3.0/downsample/downsample.0/Conv_output_0
152,             Add, 2, 1, 5_/layer3/layer3.0/conv2/Conv_output_0, 5_/layer3/layer3.0/Add_output_0
153,            Relu, 1, 1, 5_/layer3/layer3.0/Add_output_0, 5_/layer3/layer3.0/relu_1/Relu_output_0
154,            Conv, 3, 1, 5_/layer3/layer3.0/relu_1/Relu_output_0, 5_/layer3/layer3.1/conv1/Conv_output_0
155,            Relu, 1, 1, 5_/layer3/layer3.1/conv1/Conv_output_0, 5_/layer3/layer3.1/relu/Relu_output_0
156,            Conv, 3, 1, 5_/layer3/layer3.1/relu/Relu_output_0, 5_/layer3/layer3.1/conv2/Conv_output_0
157,             Add, 2, 1, 5_/layer3/layer3.1/conv2/Conv_output_0, 5_/layer3/layer3.1/Add_output_0
158,            Relu, 1, 1, 5_/layer3/layer3.1/Add_output_0, 5_/layer3/layer3.1/relu_1/Relu_output_0
159,            Conv, 3, 1, 5_/layer3/layer3.1/relu_1/Relu_output_0, 5_/neck/deblocks.3/deblocks.3.0/Conv_output_0
160,            Relu, 1, 1, 5_/neck/deblocks.3/deblocks.3.0/Conv_output_0, 5_/neck/deblocks.3/deblocks.3.2/Relu_output_0
161,          Resize, 3, 1, 5_/neck/deblocks.3/deblocks.3.2/Relu_output_0, 5_/neck/Resize_2_output_0
162,          Concat, 4, 1, 5_/neck/deblocks.0/deblocks.0.2/Relu_output_0, 5_/neck/Concat_output_0
163,            Conv, 3, 1, 5_/neck/Concat_output_0, 5_/shared_conv/conv/Conv_output_0
164,            Relu, 1, 1, 5_/shared_conv/conv/Conv_output_0, 5_/shared_conv/activate/Relu_output_0
165,            Conv, 3, 1, 5_/shared_conv/activate/Relu_output_0, 5_/task_heads.0/reg/reg.0/conv/Conv_output_0
166,            Relu, 1, 1, 5_/task_heads.0/reg/reg.0/conv/Conv_output_0, 5_/task_heads.0/reg/reg.0/activate/Relu_output_0
167,            Conv, 3, 1, 5_/task_heads.0/reg/reg.0/activate/Relu_output_0, 5_preds
168,            Conv, 3, 1, 5_/shared_conv/activate/Relu_output_0, 5_/task_heads.0/height/height.0/conv/Conv_output_0
169,            Relu, 1, 1, 5_/task_heads.0/height/height.0/conv/Conv_output_0, 5_/task_heads.0/height/height.0/activate/Relu_output_0
170,            Conv, 3, 1, 5_/task_heads.0/height/height.0/activate/Relu_output_0, 5_257
171,            Conv, 3, 1, 5_/shared_conv/activate/Relu_output_0, 5_/task_heads.0/dim/dim.0/conv/Conv_output_0
172,            Relu, 1, 1, 5_/task_heads.0/dim/dim.0/conv/Conv_output_0, 5_/task_heads.0/dim/dim.0/activate/Relu_output_0
173,            Conv, 3, 1, 5_/task_heads.0/dim/dim.0/activate/Relu_output_0, 5_261
174,            Conv, 3, 1, 5_/shared_conv/activate/Relu_output_0, 5_/task_heads.0/rot/rot.0/conv/Conv_output_0
175,            Relu, 1, 1, 5_/task_heads.0/rot/rot.0/conv/Conv_output_0, 5_/task_heads.0/rot/rot.0/activate/Relu_output_0
176,            Conv, 3, 1, 5_/task_heads.0/rot/rot.0/activate/Relu_output_0, 5_265
177,            Conv, 3, 1, 5_/shared_conv/activate/Relu_output_0, 5_/task_heads.0/vel/vel.0/conv/Conv_output_0
178,            Relu, 1, 1, 5_/task_heads.0/vel/vel.0/conv/Conv_output_0, 5_/task_heads.0/vel/vel.0/activate/Relu_output_0
179,            Conv, 3, 1, 5_/task_heads.0/vel/vel.0/activate/Relu_output_0, 5_269
180,            Conv, 3, 1, 5_/shared_conv/activate/Relu_output_0, 5_/task_heads.0/heatmap/heatmap.0/conv/Conv_output_0
181,            Relu, 1, 1, 5_/task_heads.0/heatmap/heatmap.0/conv/Conv_output_0, 5_/task_heads.0/heatmap/heatmap.0/activate/Relu_output_0
182,            Conv, 3, 1, 5_/task_heads.0/heatmap/heatmap.0/activate/Relu_output_0, 5_273

Input tensor name -  1_mlp_input 

Input tensor name -  0_img 

Input tensor name -  4_circle_map 

Input tensor name -  4_ray_map 
Output tensor name - 5_273 
Output tensor name - 5_269 
Output tensor name - 5_265 
Output tensor name - 5_261 
Output tensor name - 5_257 
Output tensor name - 5_preds 
 Graph Domain TO version : 11In TIDL_onnxRtImportInit subgraph_name=subgraph_0
Layer 0, subgraph id subgraph_0, name=5_273
Layer 1, subgraph id subgraph_0, name=5_269
Layer 2, subgraph id subgraph_0, name=5_265
Layer 3, subgraph id subgraph_0, name=5_261
Layer 4, subgraph id subgraph_0, name=5_257
Layer 5, subgraph id subgraph_0, name=5_preds
Layer 6, subgraph id subgraph_0, name=1_mlp_input
Layer 7, subgraph id subgraph_0, name=0_img
Layer 8, subgraph id subgraph_0, name=4_circle_map
Layer 9, subgraph id subgraph_0, name=4_ray_map
In TIDL_runtimesOptimizeNet: LayerIndex = 193, dataIndex = 187 
WARNING: Inner Product Layer bias 1_/context_mlp/fc1/Gemm's coeff cannot be found(or not match) in coef file, Random coeff will be generated! Only for evaluation usage! Results are all random!
WARNING: Inner Product Layer bias 1_/context_mlp/fc2/Gemm's coeff cannot be found(or not match) in coef file, Random coeff will be generated! Only for evaluation usage! Results are all random!
WARNING: Conv Layer 5_/neck/deblocks.0/deblocks.0.0/Conv's coeff cannot be found(or not match) in coef file, Random coeff will be generated! Only for evaluation usage! Results are all random!
WARNING: Conv Layer 5_/conv1/Conv's coeff cannot be found(or not match) in coef file, Random coeff will be generated! Only for evaluation usage! Results are all random!
WARNING: Conv Layer 0_/0/layer2/layer2.0/downsample/downsample.0/Conv's coeff cannot be found(or not match) in coef file, Random coeff will be generated! Only for evaluation usage! Results are all random!
WARNING: Conv Layer 0_/0/layer2/layer2.0/conv1/Conv's coeff cannot be found(or not match) in coef file, Random coeff will be generated! Only for evaluation usage! Results are all random!

 ************** Frame index 1 : Running float import ************* 
In TIDL_runtimesPostProcessNet 
In TIDL_runtimesPostProcessNet 1
In TIDL_runtimesPostProcessNet 2
In TIDL_runtimesPostProcessNet 3
Warning: FFFFFF is not a known color.
INFORMATION: [TIDL_ResizeLayer] 5_/neck/Resize Any resize ratio which is power of 2 and greater than 4 will be placed by combination of 4x4 resize layer and 2x2 resize layer. For example a 8x8 resize will be replaced by 4x4 resize followed by 2x2 resize.
INFORMATION: [TIDL_ResizeLayer] 5_/neck/Resize_1 Any resize ratio which is power of 2 and greater than 4 will be placed by combination of 4x4 resize layer and 2x2 resize layer. For example a 8x8 resize will be replaced by 4x4 resize followed by 2x2 resize.
INFORMATION: [TIDL_ResizeLayer] 5_/neck/Resize_2_TIDL_0 Any resize ratio which is power of 2 and greater than 4 will be placed by combination of 4x4 resize layer and 2x2 resize layer. For example a 8x8 resize will be replaced by 4x4 resize followed by 2x2 resize.
INFORMATION: [TIDL_ResizeLayer] 5_/neck/Resize_2 Any resize ratio which is power of 2 and greater than 4 will be placed by combination of 4x4 resize layer and 2x2 resize layer. For example a 8x8 resize will be replaced by 4x4 resize followed by 2x2 resize.
****************************************************
**          4 WARNINGS          0 ERRORS          **
****************************************************
In TIDL_runtimesPostProcessNet 4
************ in TIDL_subgraphRtCreate ************ 
 The soft limit is 2048
The hard limit is 2048
MEM: Init ... !!!
MEM: Init ... Done !!!
 0.0s:  VX_ZONE_INIT:Enabled
 0.9s:  VX_ZONE_ERROR:Enabled
 0.13s:  VX_ZONE_WARNING:Enabled
 0.4771s:  VX_ZONE_INIT:[tivxInit:190] Initialization Done !!!
TIDL_initDebugTraceParams Done 

--------------------------------------------
TIDL Memory size requiement (record wise):
MemRecNum   , Space               , Attribute   , Alignment   , Size(KBytes), BasePtr     
0           , DDR Cacheable       , Persistent  ,  128, 19.25   , 0x00000000
1           , DDR Cacheable       , Persistent  ,  128, 0.64    , 0x00000000
2           , L1D                 , Scratch     ,  128, 16.00   , 0x00000000
3           , L2                  , Scratch     ,  128, 4.00    , 0x00000000
4           , L3/MSMC             , Scratch     ,  128, 56.00   , 0x00000000
5           , DDR Cacheable       , Persistent  ,  128, 1268.38 , 0x00000000
6           , DDR Cacheable       , Scratch     ,  128, 364520.00, 0x00000000
7           , L1D                 , Scratch     ,    0, 0.00    , 0x00000000
8           , DDR Cacheable       , Scratch     ,  128, 104046.88, 0x00000000
9           , L1D                 , Scratch     ,    0, 0.00    , 0x00000000
10          , DDR Cacheable       , Persistent  ,  128, 1684.19 , 0x00000000
11          , DDR Cacheable       , Scratch     ,  128, 512.25  , 0x00000000
12          , L1D                 , Scratch     ,    0, 0.00    , 0x00000000
13          , L1D                 , Scratch     ,    0, 0.00    , 0x00000000
14          , DDR Cacheable       , Persistent  ,  128, 0.00    , 0x00000000
--------------------------------------------
Total memory size requirement (space wise):
Mem Space , Size(KBytes)
L1D       , 16.00   
L2        , 4.00    
L3/MSMC   , 56.00   
DDR Cacheable, 472051.58
--------------------------------------------
NOTE: Memory requirement in host emulation can be different from the same on EVM
      To get the actual TIDL memory requirement make sure to run on EVM with 
      debugTraceLevel = 2

--------------------------------------------
 0.736390s:  VX_ZONE_ERROR:[tivxAlgiVisionCreate:363] Calling ialg.algAlloc failed with status = -1120
 0.736463s:  VX_ZONE_ERROR:[tivxKernelTIDLCreate:926] tivxAlgiVisionCreate returned NULL
 0.736625s:  VX_ZONE_ERROR:[ownContextSendCmd:875] Command ack message returned failure cmd_status: -1
 0.736644s:  VX_ZONE_ERROR:[ownNodeKernelInit:590] Target kernel, TIVX_CMD_NODE_CREATE failed for node TIDLNode
 0.736652s:  VX_ZONE_ERROR:[ownNodeKernelInit:591] Please be sure the target callbacks have been registered for this core
 0.736661s:  VX_ZONE_ERROR:[ownNodeKernelInit:592] If the target callbacks have been registered, please ensure no errors are occurring within the create callback of this kernel
 0.736671s:  VX_ZONE_ERROR:[ownGraphNodeKernelInit:608] kernel init for node 0, kernel com.ti.tidl:4:6 ... failed !!!
 0.736680s:  VX_ZONE_ERROR:[vxVerifyGraph:2159] Node kernel init failed
 0.736688s:  VX_ZONE_ERROR:[vxVerifyGraph:2213] Graph verify failed
TIDL_RT_OVX: ERROR: Verifying TIDL graph ... Failed !!!
TIDL_RT_OVX: ERROR: Verify OpenVX graph failed
************ TIDL_subgraphRtCreate done ************ 
 *******   In TIDL_subgraphRtInvoke  ******** 
 1.634656s:  VX_ZONE_ERROR:[tivxKernelTIDLCreate:907] Network version - 0x00000000, Expected version - 0x20240401
 1.634894s:  VX_ZONE_ERROR:[ownContextSendCmd:875] Command ack message returned failure cmd_status: -1
 1.634911s:  VX_ZONE_ERROR:[ownNodeKernelInit:590] Target kernel, TIVX_CMD_NODE_CREATE failed for node TIDLNode
 1.634917s:  VX_ZONE_ERROR:[ownNodeKernelInit:591] Please be sure the target callbacks have been registered for this core
 1.634924s:  VX_ZONE_ERROR:[ownNodeKernelInit:592] If the target callbacks have been registered, please ensure no errors are occurring within the create callback of this kernel
 1.634958s:  VX_ZONE_ERROR:[ownGraphNodeKernelInit:608] kernel init for node 0, kernel com.ti.tidl:4:6 ... failed !!!
 1.634968s:  VX_ZONE_ERROR:[vxVerifyGraph:2159] Node kernel init failed
 1.634975s:  VX_ZONE_ERROR:[vxVerifyGraph:2213] Graph verify failed
 1.635011s:  VX_ZONE_ERROR:[ownGraphScheduleGraphWrapper:885] graph is not in a state required to be scheduled
 1.635018s:  VX_ZONE_ERROR:[vxProcessGraph:813] schedule graph failed
 1.635023s:  VX_ZONE_ERROR:[vxProcessGraph:818] wait graph failed
ERROR: Running TIDL graph ... Failed !!!
Sub Graph Stats 3609.000000 617023.000000 15780042465065548.000000 
*******  TIDL_subgraphRtInvoke done  ******** 

**********  Frame Index 1 : Running float inference **********
 Graph Domain TO version : 11*******   In TIDL_subgraphRtInvoke  ******** 
 1.797504s:  VX_ZONE_ERROR:[tivxKernelTIDLCreate:907] Network version - 0x00000000, Expected version - 0x20240401
 1.797718s:  VX_ZONE_ERROR:[ownContextSendCmd:875] Command ack message returned failure cmd_status: -1
 1.797730s:  VX_ZONE_ERROR:[ownNodeKernelInit:590] Target kernel, TIVX_CMD_NODE_CREATE failed for node TIDLNode
 1.797733s:  VX_ZONE_ERROR:[ownNodeKernelInit:591] Please be sure the target callbacks have been registered for this core
 1.797736s:  VX_ZONE_ERROR:[ownNodeKernelInit:592] If the target callbacks have been registered, please ensure no errors are occurring within the create callback of this kernel
 1.797741s:  VX_ZONE_ERROR:[ownGraphNodeKernelInit:608] kernel init for node 0, kernel com.ti.tidl:4:6 ... failed !!!
 1.797747s:  VX_ZONE_ERROR:[vxVerifyGraph:2159] Node kernel init failed
 1.797750s:  VX_ZONE_ERROR:[vxVerifyGraph:2213] Graph verify failed
 1.797766s:  VX_ZONE_ERROR:[ownGraphScheduleGraphWrapper:885] graph is not in a state required to be scheduled
 1.797770s:  VX_ZONE_ERROR:[vxProcessGraph:813] schedule graph failed
 1.797773s:  VX_ZONE_ERROR:[vxProcessGraph:818] wait graph failed
ERROR: Running TIDL graph ... Failed !!!
Sub Graph Stats 9927.000000 761.000000 15780042465065548.000000 
*******  TIDL_subgraphRtInvoke done  ******** 

**********  Frame Index 2 : Running fixed point mode for calibration **********
In TIDL_runtimesPostProcessNet 
In TIDL_runtimesPostProcessNet 1
In TIDL_runtimesPostProcessNet 2
In TIDL_runtimesPostProcessNet 3
Empty prototxt path, running calibration

~~~~~Running TIDL in PC emulation mode to collect Activations range for each layer~~~~~

Processing config file #0 : /home/ht/edgeai/edgeai-tidl-tools/model-artifacts/motorvis_matrix/tempDir/subgraph_0_tidl_io_.qunat_stats_config.txt 
Error at line:  1167 : in file /OpenACC/J721S2/c7x-mma-tidl/ti_dl/test/src/pc_linux/../tidl_rt.c, of function : TIDLRT_create 
Invalid Error Type!
Error at line:   527 : in file /OpenACC/J721S2/c7x-mma-tidl/ti_dl/test/src/pc_linux/../tidl_tb.c, of function : tidlMultiInstanceTest 
Invalid Error Type!

 
 
 *****************   Calibration iteration number 0 started ************************ 
 
 
 
Empty prototxt path, running calibration
^C^Z[1]   Killed                  python3 multi_input_no_post.py -c -m motorvis_matrix

[2]+  Stopped                 python3 multi_input_no_post.py -c -m motorvis_matrix

Regards,

Adam

0 Pratik Kedar over 1 year ago in reply to Adam Hua

TI__Mastermind 24041 points

Hi,

I am currently Out of Office, I will be back in the office by 24th June, please expect delay in my response.

Thank you of your kind understanding.

0 Wu han over 1 year ago in reply to Pratik Kedar

Prodigy 10 points

Hello, today is 24th June, can the above question be feedbacked today?

0 Pratik Kedar over 1 year ago in reply to Wu han

TI__Mastermind 24041 points

Hi,

We will get back to you on this thread, in coming 2 days.

Thank You

0 Pratik Kedar over 1 year ago in reply to Pratik Kedar

TI__Mastermind 24041 points

Hi Team,

Apologies for the late reply due to limited bandwidth we have made 9.2.9.0 sdk release in the meanwhile which consist of bug fixes and some improvements.

Adam will be trying the shared model import on that to validate few things, we plan to take it from there.

Thanks for your kind patience

Processors

Processors forum

TDA4VM: BEV model deploy