Forums - snpe-tensorflow-to-dlc ssd_mobilenet_v1 model conversion issue

2 posts / 0 new
Last post
snpe-tensorflow-to-dlc ssd_mobilenet_v1 model conversion issue
Join Date: 19 Sep 18
Posts: 1
Posted: Fri, 2019-02-01 13:41
I followed the SNPE  tensorflow model conversion procedure as explained here  (i also downloaded the model from the link provided here)
But i get the following issue. 
$snpe-tensorflow-to-dlc --graph ssd_mobilenet_v1_coco_2017_11_17/frozen_inference_graph.pb -i Preprocessor/sub 1,300,300,3 --out_node detection_classes --out_node detection_boxes --out_node detection_scores --dlc mobilenet_ssd.dlc --allow_unconsumed_nodes
: I tensorflow/core/platform/] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2
: F tensorflow/compiler/jit/] Check failed: it != predicate_map_.end() _SINK
Aborted (core dumped)
Here is my enviroment.
Ubuntu 14.04
python 2.7
pip 19.0
virtualenv 16.2.0
tensorflow 1.12.0


What is going wrong?



  • Up0
  • Down0
Join Date: 4 Nov 18
Posts: 68
Posted: Tue, 2019-02-05 01:58
It looks like your SNPE setup is not compatible with the latest version of TensorFlow or may be the source tree of TensorFlow is the mismatch. I have the setup using TensorFlow 1.11 on Ubuntu 16.04 and SNPE 1.21.0, which is successfully converting the model. Can you try with version 1.11 TensorFlow and clone the TensorFlow repo again. Initialize the environmental variables with TensorFlow source tree, running the below command -
source bin/ -t $TENSORFLOW_DIR
Once done, try converting the model.
  • 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.