Forums - Layer parameter value is invalid. No backend could validate Op=NonMaxSuppression

1 post / 0 new
Layer parameter value is invalid. No backend could validate Op=NonMaxSuppression
nickncc2122
Join Date: 4 Jun 24
Posts: 1
Posted: Thu, 2024-06-06 05:35

I would like to run YOLOv8 on DSP using SNPE. I have already converted the .pth model to ONNX and performed DLC quantization with the following command. 

 
my snpe version is SNPE v2.23.0.240531124219_94801
 
```
snpe-dlc-quantize  \
      --input_dlc ${output_dir}/${dataset}/${config}/ONNX-DLC/${model}.dlc \
      --input_list ${output_dir}/${dataset}/${config}/ONNX-DLC/SNPE_RAWS/raw_list.txt  \
      --use_enhanced_quantizer \
      --use_adjusted_weights_quantizer \
      --axis_quant \
      --output_dlc ${output_dir}/${dataset}/${config}/ONNX-DLC/quant_${model}.dlc \
      --enable_htp
```
 
 
However, when I attempt to run it on the DSP, I encounter an error, whereas it operates normally on the CPU.
 
`snpe-net-run --container quant_${model}.dlc --input_list rel_raw_list.txt --use_dsp --output_dir output_htp`
 
this is the error:
 
`error_code=1002; error_message=Layer parameter value is invalid. No backend could validate Op=NonMaxSuppression_318 Type=NonMaxSuppression error code=3110; error_component=Model Validation; line_no=142; thread_id=488582993144`
 
 
 
 
 
  • Up0
  • Down0

Opinions expressed in the content posted here are the personal opinions of the original authors, and do not necessarily reflect those of Qualcomm Incorporated or its subsidiaries (“Qualcomm”). The content is provided for informational purposes only and is not meant to be an endorsement or representation by Qualcomm or any other party. This site may also provide links or references to non-Qualcomm sites and resources. Qualcomm makes no representations, warranties, or other commitments whatsoever about any non-Qualcomm sites or third-party resources that may be referenced, accessible from, or linked to this site.