Forums - Coverting tensorflow model to DLC : Memory error

2 posts / 0 new
Last post
Coverting tensorflow model to DLC : Memory error
bnarendar89
Join Date: 21 Jan 19
Posts: 5
Posted: Mon, 2019-04-15 00:24

Hi ,

I want to convert the tensorflow model into DLC by using " snpe-tensorflow-to-dlc " tool . The command is as follows:

snpe-sdk$:  snpe-tensorflow-to-dlc --graph $SNPE_ROOT/models/inception_v3/tensorflow/inception_v3_2016_08_28_frozen.pb --input_dim input "1,299,299,3" --out_node "InceptionV3/Predictions/Reshape_1" --dlc /home/tcs/Downloads/Narendar/snpe-1.24.0.256/models/inception_v3/dlc/inception_v3.dlc --allow_unconsumed_nodes --verbose   

However, Im getting the following error saying as Memory Error:

2019-04-15 11:30:02,303 - 322 - INFO - INFO_TF_BUILDING_INPUT_LAYER: Building layer (INPUT) with node: input, shape [1, 299, 299, 3]
2019-04-15 11:31:14,953 - 102 - ERROR - Encountered Error: 
Traceback (most recent call last):
  File "/home/tcs/Downloads/Narendar/snpe-1.24.0.256/bin/x86_64-linux-clang/snpe-tensorflow-to-dlc", line 96, in main
    converter.convert(args.dlc, args.copyright_file, args.model_version, converter_command)
  File "/home/tcs/Downloads/Narendar/snpe-1.24.0.256/lib/python/converters/tensorflow/converter.py", line 301, in convert
    self._convert_layers()
  File "/home/tcs/Downloads/Narendar/snpe-1.24.0.256/lib/python/converters/tensorflow/converter.py", line 335, in _convert_layers
    descriptors = self._resolve_descriptors_from_nodes(graph_ops)
  File "/home/tcs/Downloads/Narendar/snpe-1.24.0.256/lib/python/converters/tensorflow/converter.py", line 471, in _resolve_descriptors_from_nodes
    resolved_descriptors = resolver.resolve_layer(graph_matcher, self._graph_helper)
  File "/home/tcs/Downloads/Narendar/snpe-1.24.0.256/lib/python/converters/tensorflow/layers/ssd.py", line 138, in resolve_layer
    matches = graph_matcher.match_sequence(ssd.nms_sequence)
  File "/home/tcs/Downloads/Narendar/snpe-1.24.0.256/lib/python/converters/tensorflow/graph_matcher.py", line 206, in match_sequence
    roots_candidate_assignments = self._find_roots_candidate_assignments(sequence.output_nodes)
  File "/home/tcs/Downloads/Narendar/snpe-1.24.0.256/lib/python/converters/tensorflow/graph_matcher.py", line 223, in _find_roots_candidate_assignments
    combinations.append(candidate)
MemoryError
 
Request your support on same..
 
Thanks,
Narendar
  • Up0
  • Down0
gesqdn-forum
Join Date: 4 Nov 18
Posts: 184
Posted: Tue, 2019-04-23 08:34

Hi bnarendar89,

This error may causes because of non-availability of enough memory in your system which occuring while loading the required data into the main memory. Before running the conversion please make sure you have enough free memory available in your system.



 

  • 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.