Part Number: TDA4VM
HI Team,
we are working on NN model and able to generate binary file.
Import traces:
------------------ Network Compiler Traces ----------------------------- successful Memory allocation INFORMATION: [TIDL_ResizeLayer] Upsample_17 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] Upsample_23 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] Upsample_29 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] Upsample_35 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 ** ****************************************************
INference traces:
ti-psdkra/tidl_j7/ti_dl/test$ ./PC_dsp_test_dl_algo.out
Processing config file #0 : testvecs/config/infer/public/onnx/tidl.txt
----------------------- TIDL Process with REF_ONLY FLOW ------------------------
- 0 . .. T 3024.47 .... ..... ... A : 0, 0.0000, 0.0000, 103275 .... .....p
As we are using tidl-RT executioner to get output is of binary file (.bin) extension.
what format will this data be stored in the binary...?
Is there supported ways you can suggest to read these binary inference output (.bin) into Opencv to read data in some matrix for example cv::mat format...?
For example :
This is a blocker and high priority for us...!
Kindly do the needful in this regard.
Thanks & Regards,
Pawamana Tanjore.