Forums - yolov3 pb to dlc problem

2 posts / 0 new
or Register
Last post
yolov3 pb to dlc problem
frank.huang_1
Join Date: 22 Jul 20
Posts: 7
Posted: Mon, 2020-08-17 19:15
Top

Hi, I'm trying to conver yolov3 pb file to dlc file through command  below:

snpe-tensorflow-to-dlc --input_network pb_path \

           --input_dim input/input_data 1, 416,416,3 \
           --out_node pred_sbbox/concat_2   \
           --out_node pred_mbbox/concat_2 \
           --out_node pred_lbbox/concat_2    \
           --output_path dlc_name                   \
           --allow_unconsumed_nodes
 
however, after running for a moment, I get this error:
 
I search on the internet and find other pople also have this problem, but they didn't mention how to solve it, 
Can anyone give me some suggestion to solve this kind of problem?
Any advice will be appreciatedm thanks.
  • Up0
  • Down0
Re: yolov3 pb to dlc problem
frank.huang_1
Join Date: 22 Jul 20
Posts: 7
Posted: Mon, 2020-08-17 19:27
Top

The error message shows below:

2020-08-18 10:25:45,827 - 166 - ERROR - %
Traceback (most recent call last):
  File "/media/ubuntu/5a65e304-34b4-4b32-ab2d-31a913d37c99/snpe-1.41.0.2173/lib/python/qti/aisw/converters/tensorflow/tf_to_ir.py", line 525, in _create_layer
    layer_builder.build_layer(self.graph, self._context, descriptor, inputs, outputs)
  File "/media/ubuntu/5a65e304-34b4-4b32-ab2d-31a913d37c99/snpe-1.41.0.2173/lib/python/qti/aisw/converters/tensorflow/layers/resize.py", line 243, in build_layer
    input_name = self.get_input_name(converter_context, descriptor, input_descriptors)
  File "/media/ubuntu/5a65e304-34b4-4b32-ab2d-31a913d37c99/snpe-1.41.0.2173/lib/python/qti/aisw/converters/tensorflow/common.py", line 163, in get_input_name
    input_descriptors[0].layer_type, 1, len(input_descriptors)
qti.aisw.converters.tensorflow.util.ConverterError: ERROR_TF_LAYER_INPUT_COUNT_ERROR: Layer RELU expects 1 input(s), actual 2
2020-08-18 10:25:45,851 - 166 - ERROR - Conversion failed: build_layer failed with Exception ERROR_TF_LAYER_INPUT_COUNT_ERROR: Layer RELU expects 1 input(s), actual 2 in layer ResizeLayerBuilder
 
  • Up0
  • Down0
or Register

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.