J721EXCPXEVM: TDA4VMXEVM: [TIDL] import onnx model to tidl model error.

Nikita Verbitsky

Part Number: J721EXCPXEVM

Hi,

I am trying to import BiSeNet model, but even after converting them from PyTorch to ONNX (opset_ver = 11 , there is no way to downgrade the opset_ver of the model below 11), importing them using the import tool gives me some form of error. Error log has shown below.
1) What is the reason for the import error shown in the log? How can I debug it?
2) Why did the "find Indata for data ID" lines appear when importing? ID refers to a layer? If so, how can I make a comparison with the architecture in the ONNX model?
3) Is the error related to the lack of Sigmoid layer in TIDL?
4) Is there any documentation for implementing custom layers for TIDL, other than the source code example at path psdk_rtos_auto_j7_07_00_00_11/tidl_j7_01_02_00_09/ti_dl/custom?
5) How do I update the SDK (PSDKLA & PSDKLA) to 20th Nov release? Should I delete the SDK folders and perform a clean install?

Error log:
nikita@VERBITSKY-N:~/psdk_rtos_auto_j7_07_00_00_11/tidl_j7_01_02_00_09/ti_dl/utils/tidlModelImport$ ./out/tidl_model_import.out /home/nikita/Work/TDA4/tidl_bisenet/tidl_configs/tidl_import_bisenet.txt

ONNX Model (Proto) File : /home/nikita/Work/TDA4/tidl_bisenet/snapshot/qat_model4.onnx

TIDL Network File : /home/nikita/Work/TDA4/tidl_bisenet/tidl_models/onnx/tidl_net_qat_model4.bin

TIDL IO Info File : /home/nikita/Work/TDA4/tidl_bisenet/tidl_models/onnx/tidl_io_qat_model4_

Current ONNX OpSet Version : 11

ONNX operator Sigmoid is not suported now.. By passing

ONNX operator Sigmoid is not suported now.. By passing

ONNX operator Sigmoid is not suported now.. By passing

Could not find Indata for data ID 36

Could not find Indata for data ID 47

Could not find Indata for data ID 60

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

Processing config file #0 : /home/nikita/Work/TDA4/tidl_bisenet/tidl_models/onnx/tidl_import_bisenet.txt.qunat_stats_config.txt

Algorithm Init failed with error number: -1000

Error at line: 705 : in file src/tidl_tb.c, of function : tidlMultiInstanceTest

Error Type: TIDL_E_CONV_INVALID_INPUT_WIDTH

****************************************************

** All the Tensor Dimensions has to be greater then Zero

** DIM Error - For Tensor 45, Dim 2 is 0

****************************************************

Segmentation fault (core dumped)

WARNING: [TIDL_ResizeLayer] Resize_113 Resize kernel with non-power of 2 resize ratio is not optimal.

WARNING: [TIDL_ResizeLayer] Resize_142 Resize kernel with non-power of 2 resize ratio is not optimal.

WARNING: [TIDL_ResizeLayer] Resize_174 Resize kernel with non-power of 2 resize ratio is not optimal.

INFORMATION : 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

INFORMATION : Example a 8x8 resize will be replaced by 4x4 resize followed by 2x2 resize

INFORMATION : 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

INFORMATION : Example a 8x8 resize will be replaced by 4x4 resize followed by 2x2 resize

ERROR : [TIDL_E_QUANT_STATS_NOT_AVAILABLE] tidl_quant_stats_tool.out fails to collect dynamic range. Please look into quant stats log. This model will get fault on target.

WARNING: [TIDL_E_DATAFLOW_INFO_NULL] ti_cnnperfsim.out fails to allocate memory in MSMC. Please look into perfsim log. This model can only be used on PC emulation, it will get fault on target.

****************************************************

** 4 WARNINGS 1 ERRORS **

****************************************************
File import model config: tidl_import_bisenet.txt
modelType = 2
inputNetFile = "/home/nikita/Work/TDA4/tidl_bisenet/snapshot/qat_model4.onnx"
outputNetFile = "/home/nikita/Work/TDA4/tidl_bisenet/tidl_models/onnx/tidl_net_qat_model4.bin"
outputParamsFile = "/home/nikita/Work/TDA4/tidl_bisenet/tidl_models/onnx/tidl_io_qat_model4_"
inWidth = 640
inHeight = 320
inNumChannels = 3
inDataFormat = 0
inData = "/home/nikita/Work/TDA4/tidl_bisenet/tidl_configs/val.txt"
postProcType = 3
perfSimConfig = ../../test/testvecs/config/import/device_config.cfg
File tidl_import_bisenet.txt.qunat_stats_config.txt:
inFileFormat = 2
numFrames = 245
postProcType = 3
postProcDataId = 0
quantRangeUpdateFactor = -1.000000
inData = /home/nikita/Work/TDA4/tidl_bisenet/tidl_configs/val.txt
outData = "/home/nikita/Work/TDA4/tidl_bisenet/tidl_models/onnx/tidl_import_bisenet.txt_stats_tool_out.bin"
netBinFile = /home/nikita/Work/TDA4/tidl_bisenet/tidl_models/onnx/tidl_net_qat_model4.bin
ioConfigFile = /home/nikita/Work/TDA4/tidl_bisenet/tidl_models/onnx/tidl_io_qat_model4_1.bin
flowCtrl = 3
writeTraceLevel = 0
debugTraceLevel = 0
traceDumpBaseName = "/home/nikita/Work/TDA4/tidl_bisenet/tidl_models/onnx/tidl_import_bisenet.txt

over 4 years ago

0 kumar.desappan over 4 years ago

TI__Mastermind 22145 points

Hi ,

This model is failing because of the below un supported layer.

ONNX operator Sigmoid is not suported now.. By passing

Sigmoid layer is not supported in 7.0 release.

Please migrate to SDK 7.1 for sigmoid layer support.

0 Nikita Verbitsky over 4 years ago in reply to kumar.desappan

Prodigy 125 points

Hi,

After upgrading to SDK 7.1, I no longer get warnings about missing layers, such as Sigmoid. Thanks.

Unfortunately, when importing the model (trained in QAT mode) from ONNX to BIN format, the model will have the Resize 46 layer ("DIM Error - For Tensor 46, Dim 2 is 0"). What can this problem be related to?

Log:

ONNX Model (Proto) File  : /home/nikita/tidlBisenet/snapshot/qat_model0t7.onnx  
TIDL Network File      : /home/nikita/tidlBisenet/tidl_models/onnx/tidl_net_qat_model0t7.bin  
TIDL IO Info File      : /home/nikita/tidlBisenet/tidl_models/onnx/tidl_io_qat_model0t7_  
Current ONNX OpSet Version   : 11  
Running tidl_optimizeNet
Completed tidl_optimizeNet
Num of Layer Detected :  69
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
  Num|TIDL Layer Name               |Out Data Name                                     |Group |#Ins  |#Outs |Inbuf Ids                       |Outbuf Id |In NCHW                             |Out NCHW                            |MACS       |
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
    0|TIDL_DataLayer                |x.1_original                                      |     0|    -1|     1|  x   x   x   x   x   x   x   x |  0       |       0        0        0        0 |       1        3      320      640 |         0 |
    1|TIDL_BatchNormLayer           |x.1                                               |     0|     1|     1|  0   x   x   x   x   x   x   x |  1       |       1        3      320      640 |       1        3      320      640 |    614400 |
    2|TIDL_BatchNormLayer           |319                                               |     0|     1|     1|  1   x   x   x   x   x   x   x |  2       |       1        3      320      640 |       1        3      320      640 |    614400 |
    3|TIDL_ConvolutionLayer         |354                                               |     0|     1|     1|  1   x   x   x   x   x   x   x |  3       |       1        3      320      640 |       1       64      160      320 | 488243200 |
    4|TIDL_ConvolutionLayer         |326                                               |     0|     1|     1|  2   x   x   x   x   x   x   x |  4       |       1        3      320      640 |       1       64      160      320 | 488243200 |
    5|TIDL_PoolingLayer             |355                                               |     0|     1|     1|  3   x   x   x   x   x   x   x |  5       |       1       64      160      320 |       1       64       80      160 |   7372800 |
    6|TIDL_ConvolutionLayer         |333                                               |     0|     1|     1|  4   x   x   x   x   x   x   x |  6       |       1       64      160      320 |       1       64       80      160 | 473497600 |
    7|TIDL_BatchNormLayer           |360                                               |     0|     1|     1|  5   x   x   x   x   x   x   x |  7       |       1       64       80      160 |       1       64       80      160 |    819200 |
    8|TIDL_ConvolutionLayer         |367                                               |     0|     1|     1|  7   x   x   x   x   x   x   x |  8       |       1       64       80      160 |       1       64       80      160 | 473497600 |
    9|TIDL_ConvolutionLayer         |340                                               |     0|     1|     1|  6   x   x   x   x   x   x   x |  9       |       1       64       80      160 |       1       64       40       80 | 118374400 |
   10|TIDL_ConvolutionLayer         |374                                               |     0|     1|     1|  8   x   x   x   x   x   x   x | 10       |       1       64       80      160 |       1       64       80      160 | 473497600 |
   11|TIDL_ConvolutionLayer         |347                                               |     0|     1|     1|  9   x   x   x   x   x   x   x | 11       |       1       64       40       80 |       1      128       40       80 |  27033600 |
   12|TIDL_EltWiseLayer             |380                                               |     0|     2|     1| 10   7   x   x   x   x   x   x | 12       |       1       64       80      160 |       1       64       80      160 |   1638400 |
   13|TIDL_ConvolutionLayer         |387                                               |     0|     1|     1| 12   x   x   x   x   x   x   x | 13       |       1       64       80      160 |       1       64       80      160 | 473497600 |
   14|TIDL_ConvolutionLayer         |394                                               |     0|     1|     1| 13   x   x   x   x   x   x   x | 14       |       1       64       80      160 |       1       64       80      160 | 473497600 |
   15|TIDL_EltWiseLayer             |400                                               |     0|     2|     1| 14  12   x   x   x   x   x   x | 15       |       1       64       80      160 |       1       64       80      160 |   1638400 |
   16|TIDL_ConvolutionLayer         |407                                               |     0|     1|     1| 15   x   x   x   x   x   x   x | 16       |       1       64       80      160 |       1      128       40       80 | 236748800 |
   17|TIDL_ConvolutionLayer         |421                                               |     0|     1|     1| 15   x   x   x   x   x   x   x | 17       |       1       64       80      160 |       1      128       40       80 |  27033600 |
   18|TIDL_ConvolutionLayer         |414                                               |     0|     1|     1| 16   x   x   x   x   x   x   x | 18       |       1      128       40       80 |       1      128       40       80 | 472678400 |
   19|TIDL_EltWiseLayer             |427                                               |     0|     2|     1| 18  17   x   x   x   x   x   x | 19       |       1      128       40       80 |       1      128       40       80 |    819200 |
   20|TIDL_ConvolutionLayer         |434                                               |     0|     1|     1| 19   x   x   x   x   x   x   x | 20       |       1      128       40       80 |       1      128       40       80 | 472678400 |
   21|TIDL_ConvolutionLayer         |441                                               |     0|     1|     1| 20   x   x   x   x   x   x   x | 21       |       1      128       40       80 |       1      128       40       80 | 472678400 |
   22|TIDL_EltWiseLayer             |447                                               |     0|     2|     1| 21  19   x   x   x   x   x   x | 22       |       1      128       40       80 |       1      128       40       80 |    819200 |
   23|TIDL_ConvolutionLayer         |454                                               |     0|     1|     1| 22   x   x   x   x   x   x   x | 23       |       1      128       40       80 |       1      256       20       40 | 236339200 |
   24|TIDL_ConvolutionLayer         |468                                               |     0|     1|     1| 22   x   x   x   x   x   x   x | 24       |       1      128       40       80 |       1      256       20       40 |  26624000 |
   25|TIDL_ConvolutionLayer         |461                                               |     0|     1|     1| 23   x   x   x   x   x   x   x | 25       |       1      256       20       40 |       1      256       20       40 | 472268800 |
   26|TIDL_EltWiseLayer             |474                                               |     0|     2|     1| 25  24   x   x   x   x   x   x | 26       |       1      256       20       40 |       1      256       20       40 |    409600 |
   27|TIDL_ConvolutionLayer         |481                                               |     0|     1|     1| 26   x   x   x   x   x   x   x | 27       |       1      256       20       40 |       1      256       20       40 | 472268800 |
   28|TIDL_ConvolutionLayer         |488                                               |     0|     1|     1| 27   x   x   x   x   x   x   x | 28       |       1      256       20       40 |       1      256       20       40 | 472268800 |
   29|TIDL_EltWiseLayer             |494                                               |     0|     2|     1| 28  26   x   x   x   x   x   x | 29       |       1      256       20       40 |       1      256       20       40 |    409600 |
   30|TIDL_ConvolutionLayer         |515                                               |     0|     1|     1| 29   x   x   x   x   x   x   x | 30       |       1      256       20       40 |       1      512       10       20 |  26419200 |
   31|TIDL_ConvolutionLayer         |501                                               |     0|     1|     1| 29   x   x   x   x   x   x   x | 31       |       1      256       20       40 |       1      512       10       20 | 236134400 |
   32|TIDL_ConvolutionLayer         |619                                               |     0|     1|     1| 29   x   x   x   x   x   x   x | 32       |       1      256       20       40 |       1      128       20       40 | 236134400 |
   33|TIDL_PoolingLayer             |620                                               |     0|     1|     1| 32   x   x   x   x   x   x   x | 33       |       1      128       20       40 |       1      128        1        1 |       128 |
   34|TIDL_ConvolutionLayer         |508                                               |     0|     1|     1| 31   x   x   x   x   x   x   x | 34       |       1      512       10       20 |       1      512       10       20 | 472064000 |
   35|TIDL_ConvolutionLayer         |627                                               |     0|     1|     1| 33   x   x   x   x   x   x   x | 35       |       1      128        1        1 |       1      128        1        1 |     16640 |
   36|TIDL_EltWiseLayer             |521                                               |     0|     2|     1| 34  30   x   x   x   x   x   x | 36       |       1      512       10       20 |       1      512       10       20 |    204800 |
   37|TIDL_BatchNormLayer           |628                                               |     0|     1|     1| 35   x   x   x   x   x   x   x | 37       |       1      128        1        1 |       1      128        1        1 |       128 |
   38|TIDL_ConvolutionLayer         |528                                               |     0|     1|     1| 36   x   x   x   x   x   x   x | 38       |       1      512       10       20 |       1      512       10       20 | 472064000 |
   39|TIDL_EltWiseLayer             |629                                               |     0|     2|     1| 32  37   x   x   x   x   x   x | 39       |       1      128       20       40 |       1      128       20       40 |    102400 |
   40|TIDL_ConvolutionLayer         |535                                               |     0|     1|     1| 38   x   x   x   x   x   x   x | 40       |       1      512       10       20 |       1      512       10       20 | 472064000 |
   41|TIDL_EltWiseLayer             |541                                               |     0|     2|     1| 40  36   x   x   x   x   x   x | 41       |       1      512       10       20 |       1      512       10       20 |    204800 |
   42|TIDL_PoolingLayer             |542                                               |     0|     1|     1| 41   x   x   x   x   x   x   x | 42       |       1      512       10       20 |       1      512        1        1 |       512 |
   43|TIDL_ConvolutionLayer         |575                                               |     0|     1|     1| 41   x   x   x   x   x   x   x | 43       |       1      512       10       20 |       1      128       10       20 | 118016000 |
   44|TIDL_ConvolutionLayer         |549                                               |     0|     1|     1| 42   x   x   x   x   x   x   x | 44       |       1      512        1        1 |       1      128        1        1 |     65792 |
   45|TIDL_PoolingLayer             |576                                               |     0|     1|     1| 43   x   x   x   x   x   x   x | 45       |       1      128       10       20 |       1      128        1        1 |       128 |
   46|TIDL_ResizeLayer              |568                                               |     0|     1|     1| 44   x   x   x   x   x   x   x | 46       |       1      128        1        1 |       1      128        0        0 |         0 |
   47|TIDL_ConvolutionLayer         |583                                               |     0|     1|     1| 45   x   x   x   x   x   x   x | 47       |       1      128        1        1 |       1      128        1        1 |     16640 |
   48|TIDL_BatchNormLayer           |584                                               |     0|     1|     1| 47   x   x   x   x   x   x   x | 48       |       1      128        1        1 |       1      128        1        1 |       128 |
   49|TIDL_EltWiseLayer             |585                                               |     0|     2|     1| 43  48   x   x   x   x   x   x | 49       |       1      128       10       20 |       1      128       10       20 |     25600 |
   50|TIDL_EltWiseLayer             |586                                               |     0|     2|     1| 49  46   x   x   x   x   x   x | 50       |       1      128       10       20 |       1      128       10       20 |     25600 |
   51|TIDL_ResizeLayer              |605                                               |     0|     1|     1| 50   x   x   x   x   x   x   x | 51       |       1      128       10       20 |       1      128        0        0 |         0 |
   52|TIDL_ConvolutionLayer         |612                                               |     0|     1|     1| 51   x   x   x   x   x   x   x | 52       |       1      128        0        0 |       1      128        0        0 |         0 |
   53|TIDL_EltWiseLayer             |630                                               |     0|     2|     1| 39  52   x   x   x   x   x   x | 53       |       1      128       20       40 |       1      128       20       40 |    102400 |
   54|TIDL_ResizeLayer              |649                                               |     0|     1|     1| 53   x   x   x   x   x   x   x | 54       |       1      128       20       40 |       1      128        0        0 |         0 |
   55|TIDL_ConvolutionLayer         |656                                               |     0|     1|     1| 54   x   x   x   x   x   x   x | 55       |       1      128        0        0 |       1      128        0        0 |         0 |
   56|TIDL_ConcatLayer              |657                                               |     0|     2|     1| 11  55   x   x   x   x   x   x | 56       |       1      128       40       80 |       1      256       40       80 |    819200 |
   57|TIDL_ConvolutionLayer         |664                                               |     0|     1|     1| 56   x   x   x   x   x   x   x | 57       |       1      256       40       80 |       1      256       40       80 | 211353600 |
   58|TIDL_PoolingLayer             |665                                               |     0|     1|     1| 57   x   x   x   x   x   x   x | 58       |       1      256       40       80 |       1      256        1        1 |       256 |
   59|TIDL_ConvolutionLayer         |671                                               |     0|     1|     1| 58   x   x   x   x   x   x   x | 59       |       1      256        1        1 |       1       64        1        1 |     16448 |
   60|TIDL_ConvolutionLayer         |677                                               |     0|     1|     1| 59   x   x   x   x   x   x   x | 60       |       1       64        1        1 |       1      256        1        1 |     16640 |
   61|TIDL_BatchNormLayer           |678                                               |     0|     1|     1| 60   x   x   x   x   x   x   x | 61       |       1      256        1        1 |       1      256        1        1 |       256 |
   62|TIDL_EltWiseLayer             |679                                               |     0|     2|     1| 57  61   x   x   x   x   x   x | 62       |       1      256       40       80 |       1      256       40       80 |    819200 |
   63|TIDL_EltWiseLayer             |680                                               |     0|     2|     1| 57  62   x   x   x   x   x   x | 63       |       1      256       40       80 |       1      256       40       80 |    819200 |
   64|TIDL_ConvolutionLayer         |687                                               |     0|     1|     1| 63   x   x   x   x   x   x   x | 64       |       1      256       40       80 |       1       64       40       80 | 472268800 |
   65|TIDL_ConvolutionLayer         |693                                               |     0|     1|     1| 64   x   x   x   x   x   x   x | 65       |       1       64       40       80 |       1        8       40       80 |   1664000 |
   66|TIDL_ResizeLayer              |703_TIDL_0                                        |     0|     1|     1| 65   x   x   x   x   x   x   x | 66       |       1        8       40       80 |       1        8      160      320 |         0 |
   67|TIDL_ResizeLayer              |703                                               |     0|     1|     1| 66   x   x   x   x   x   x   x | 67       |       1        8      160      320 |       1        8      320      640 |   6553600 |
   68|TIDL_DataLayer                |703                                               |     0|     1|    -1| 67   x   x   x   x   x   x   x |  0       |       1        8      320      640 |       0        0        0        0 |         0 |
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Total Giga Macs : 9.5941
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

~~~~~Running TIDL in PC emulation mode to collect Activations range for each layer~~~~~
cd /home/nikita/ti-processor-sdk-rtos-j721e-evm-07_01_00_11/tidl_j7_01_03_00_11/ti_dl/test && ./PC_dsp_test_dl_algo.out s:/home/nikita/tidlBisenet/tidl_models/onnx/tidl_import_bisenet.txt.qunat_stats_config.txt

Processing config file #0 : /home/nikita/tidlBisenet/tidl_models/onnx/tidl_import_bisenet.txt.qunat_stats_config.txt

 Instance created for  /home/nikita/tidlBisenet/tidl_models/onnx/tidl_import_bisenet.txt.qunat_stats_config.txt
 ----------------------- TIDL Process with REF_ONLY FLOW ------------------------

#    0 . .. T    4512.07  .... ..... ... .... .....
#    1 . .. T    4506.67  .... ..... ... .... .....
#    2 . .. T    4512.33  .... ..... ... .... .....
#    3 . .. T    4505.75  .... ..... ... .... .....
#    4 . .. T    4503.66  .... ..... ... .... .....
#    5 . .. T    4501.52  .... ..... ... .... .....****************************************************
**   All the Tensor Dimensions has to be greater then Zero
**   DIM Error - For Tensor 46, Dim 2 is 0
****************************************************
/home/nikita/ti-processor-sdk-rtos-j721e-evm-07_01_00_11/tidl_j7_01_03_00_11/ti_dl/utils/tidlModelGraphviz/out/tidl_graphVisualiser.out /home/nikita/tidlBisenet/tidl_models/onnx/tidl_net_qat_model0t7.bin
****************************************************
**               TIDL Model Checker               **
****************************************************
SUGGESTION: [TIDL_ResizeLayer] Resize_113 Resize kernel with non-power of 2 resize ratio is not optimal.
SUGGESTION: [TIDL_ResizeLayer] Resize_142 Resize kernel with non-power of 2 resize ratio is not optimal.
SUGGESTION: [TIDL_ResizeLayer] Resize_174 Resize kernel with non-power of 2 resize ratio is not optimal.
INFORMATION: [TIDL_ResizeLayer] Resize_196_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] Resize_196 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.
WARNING: [TIDL_E_DATAFLOW_INFO_NULL] ti_cnnperfsim.out fails to allocate memory in MSMC. Please look into perfsim log. This model can only be used on PC emulation, it will get fault on target.
****************************************************
**          6 WARNINGS          0 ERRORS          **
****************************************************

Import config is:

modelType          = 2
inputNetFile       = "/home/nikita/tidlBisenet/snapshot/qat_model0t7.onnx"
outputNetFile      = "/home/nikita/tidlBisenet/tidl_models/onnx/tidl_net_qat_model0t7.bin"
outputParamsFile   = "/home/nikita/tidlBisenet/tidl_models/onnx/tidl_io_qat_model0t7_"
inWidth            = 640
inHeight           = 320
inNumChannels      = 3
inDataFormat       = 0
inData  =   "/home/nikita/tidlBisenet/tidl_configs/val.txt"
postProcType       = 3
perfSimConfig = ../../test/testvecs/config/import/device_config.cfg
quantizationStyle = 3
#foldPreBnConv2D = 0
#calibrationOption = 0
inDataNorm = 1
inMean = 255.0 255.0 255.0
inScale = 0.5 0.5 0.5
resizeWidth = 640
resizeHeight = 320
debugTraceLevel = 1
writeTraceLevel = 0

ImagesArch.zip

0 Nikita Verbitsky over 4 years ago in reply to Nikita Verbitsky

Prodigy 125 points

The problem was in TIDL_ResizeLayer and scale factor that was not power of 2 (width * height: 1x1 -> 10x20). In table "Core Layers/Operators Mapping & Notes" from https://software-dl.ti.com/jacinto7/esd/processor-sdk-rtos-jacinto7/latest/exports/docs/tidl_j7_01_03_00_11/ti_dl/docs/user_guide_html/md_tidl_layers_info.html, said that "TIDL_ResizeLayer only support Power of 2 and symmetric resize".

Processors

Processors forum

J721EXCPXEVM: TDA4VMXEVM: [TIDL] import onnx model to tidl model error.