Forums - Errow in converting ssd_Mobilenet model tensorflow pb file to dlc file.

7 posts / 0 new

or Register

Last post

Errow in converting ssd_Mobilenet model tensorflow pb file to dlc file.

harshiddh.mania

Join Date: 4 Jan 18

Posts: 4

Posted: Thu, 2018-02-15 16:11

Top

Hi,

I am trying to convert the ssd_mobilenet model Tensorflow .pb file into the SNPE dlc file, but i am getting error message. I am trying with below command

snpe-tensorflow-to-dlc --graph /media/sf_DL_ML/tensorflow/models/research/object_detection/ssd_mobilenet_v1_coco_2017_11_17/frozen_inference_graph.pb --input_dim image_tensor 300,300,300,3 --dlc /media/sf_DL_ML/tensorflow/models/research/object_detection/ssd_mobilenet_v1_coco_2017_11_17/ssd_mobilenet_v1.dlc --out_node detection_boxes --out_node detection_scores --out_node detection_classes --out_node num_detections --allow_unconsumed_nodes

Error Logs:

snpe-1.10.1$ snpe-tensorflow-to-dlc --graph /media/sf_DL_ML/tensorflow/models/research/object_detection/ssd_mobilenet_v1_coco_2017_11_17/frozen_inference_graph.pb --input_dim image_tensor 300,300,300,3 --dlc /media/sf_DL_ML/tensorflow/models/research/object_detection/ssd_mobilenet_v1_coco_2017_11_17/ssd_mobilenet_v1.dlc --out_node detection_boxes --out_node detection_scores --out_node detection_classes --out_node num_detections --allow_unconsumed_nodes
2018-02-15 15:33:10.914672: I tensorflow/core/platform/cpu_feature_guard.cc:137] Your CPU supports instructions that this TensorFlow binary was not compiled to use: SSE4.1 SSE4.2 AVX AVX2
2018-02-15 15:33:16.434007: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5478] has already been set.
2018-02-15 15:33:16.434145: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5478] has already been set.
2018-02-15 15:33:16.434245: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5478] has already been set.
2018-02-15 15:33:16.434304: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5478] has already been set.
2018-02-15 15:33:16.434359: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5478] has already been set.
2018-02-15 15:33:16.434416: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5478] has already been set.
2018-02-15 15:33:16.434471: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5478] has already been set.
2018-02-15 15:33:16.434525: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5478] has already been set.
2018-02-15 15:33:16.434579: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5478] has already been set.
2018-02-15 15:33:16.434633: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5472] has already been set.
2018-02-15 15:33:16.434678: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5473] has already been set.
2018-02-15 15:33:16.434736: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5490] has already been set.
2018-02-15 15:33:16.434782: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5489] has already been set.
2018-02-15 15:33:16.434823: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5467] has already been set.
2018-02-15 15:33:16.434865: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5483] has already been set.
2018-02-15 15:33:16.434907: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5471] has already been set.
2018-02-15 15:33:16.434948: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5470] has already been set.
2018-02-15 15:33:16.434990: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5472] has already been set.
2018-02-15 15:33:16.435031: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5473] has already been set.
2018-02-15 15:33:16.435072: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5476] has already been set.
2018-02-15 15:33:16.435113: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5469] has already been set.
2018-02-15 15:33:16.435164: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5490] has already been set.
2018-02-15 15:33:16.435333: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5489] has already been set.
2018-02-15 15:33:16.435385: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5467] has already been set.
2018-02-15 15:33:16.435429: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5483] has already been set.
2018-02-15 15:33:16.435471: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5471] has already been set.
2018-02-15 15:33:16.435511: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5470] has already been set.
2018-02-15 15:33:16.435553: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5472] has already been set.
2018-02-15 15:33:16.435593: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5473] has already been set.
2018-02-15 15:33:16.435634: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5476] has already been set.
2018-02-15 15:33:16.435676: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5469] has already been set.
2018-02-15 15:33:16.435716: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5490] has already been set.
2018-02-15 15:33:16.435757: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5489] has already been set.
2018-02-15 15:33:16.435797: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5467] has already been set.
2018-02-15 15:33:16.435827: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5483] has already been set.
2018-02-15 15:33:16.435836: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5471] has already been set.
2018-02-15 15:33:16.435843: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5470] has already been set.
2018-02-15 15:33:16.435851: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5472] has already been set.
2018-02-15 15:33:16.435858: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5473] has already been set.
2018-02-15 15:33:16.435866: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5476] has already been set.
2018-02-15 15:33:16.435874: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5469] has already been set.
2018-02-15 15:33:16.435882: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5490] has already been set.
2018-02-15 15:33:16.435889: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5489] has already been set.
2018-02-15 15:33:16.435896: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5467] has already been set.
2018-02-15 15:33:16.435904: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5483] has already been set.
2018-02-15 15:33:16.435911: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5471] has already been set.
2018-02-15 15:33:16.435918: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5470] has already been set.
2018-02-15 15:33:16.435926: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5472] has already been set.
2018-02-15 15:33:16.435934: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5473] has already been set.
2018-02-15 15:33:16.435941: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5476] has already been set.
2018-02-15 15:33:16.435948: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5469] has already been set.
2018-02-15 15:33:16.435955: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5490] has already been set.
2018-02-15 15:33:16.435962: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5489] has already been set.
2018-02-15 15:33:16.435970: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5467] has already been set.
2018-02-15 15:33:16.435977: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5483] has already been set.
2018-02-15 15:33:16.435984: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5471] has already been set.
2018-02-15 15:33:16.435991: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5470] has already been set.
2018-02-15 15:33:16.435998: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5472] has already been set.
2018-02-15 15:33:16.436005: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5473] has already been set.
2018-02-15 15:33:16.436012: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5476] has already been set.
2018-02-15 15:33:16.436019: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5469] has already been set.
2018-02-15 15:33:16.436027: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5490] has already been set.
2018-02-15 15:33:16.436034: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5489] has already been set.
2018-02-15 15:33:16.436041: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5467] has already been set.
2018-02-15 15:33:16.436048: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5483] has already been set.
2018-02-15 15:33:16.436055: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5471] has already been set.
2018-02-15 15:33:16.436062: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5470] has already been set.
2018-02-15 15:33:16.436069: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5472] has already been set.
2018-02-15 15:33:16.436076: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5473] has already been set.
2018-02-15 15:33:16.436084: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5476] has already been set.
2018-02-15 15:33:16.436091: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5469] has already been set.
2018-02-15 15:33:16.436098: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5490] has already been set.
2018-02-15 15:33:16.436105: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5489] has already been set.
2018-02-15 15:33:16.436112: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5467] has already been set.
2018-02-15 15:33:16.436119: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5483] has already been set.
2018-02-15 15:33:16.436126: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5471] has already been set.
2018-02-15 15:33:16.436133: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5470] has already been set.
2018-02-15 15:33:16.436140: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5472] has already been set.
2018-02-15 15:33:16.436147: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5473] has already been set.
2018-02-15 15:33:16.436154: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5476] has already been set.
2018-02-15 15:33:16.436161: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5469] has already been set.
2018-02-15 15:33:16.436168: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5490] has already been set.
2018-02-15 15:33:16.436189: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5489] has already been set.
2018-02-15 15:33:16.436198: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5467] has already been set.
2018-02-15 15:33:16.436206: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5483] has already been set.
2018-02-15 15:33:16.436213: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5471] has already been set.
2018-02-15 15:33:16.436221: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5470] has already been set.
2018-02-15 15:33:16.436228: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5472] has already been set.
2018-02-15 15:33:16.436235: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5473] has already been set.
2018-02-15 15:33:16.436244: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5476] has already been set.
2018-02-15 15:33:16.436251: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5469] has already been set.
2018-02-15 15:33:16.436258: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5490] has already been set.
2018-02-15 15:33:16.436265: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5489] has already been set.
2018-02-15 15:33:16.436273: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5467] has already been set.
2018-02-15 15:33:16.436280: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5483] has already been set.
2018-02-15 15:33:16.436287: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5471] has already been set.
2018-02-15 15:33:16.436294: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5470] has already been set.
2018-02-15 15:33:16.436301: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5469] has already been set.
2018-02-15 15:33:16.705905: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5481] has already been set.
2018-02-15 15:33:16.706023: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5480] has already been set.
2018-02-15 15:33:16.706069: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5479] has already been set.
2018-02-15 15:33:16.706114: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5487] has already been set.
2018-02-15 15:33:16.708588: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5486] has already been set.
2018-02-15 15:33:16.708896: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5481] has already been set.
2018-02-15 15:33:16.708940: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5480] has already been set.
2018-02-15 15:33:16.708984: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5479] has already been set.
2018-02-15 15:33:16.709025: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5487] has already been set.
2018-02-15 15:33:16.709068: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5486] has already been set.
2018-02-15 15:33:16.709368: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5481] has already been set.
2018-02-15 15:33:16.709413: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5480] has already been set.
2018-02-15 15:33:16.709453: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5479] has already been set.
2018-02-15 15:33:16.709494: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5487] has already been set.
2018-02-15 15:33:16.709569: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5486] has already been set.
2018-02-15 15:33:16.709774: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5481] has already been set.
2018-02-15 15:33:16.709818: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5480] has already been set.
2018-02-15 15:33:16.709859: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5479] has already been set.
2018-02-15 15:33:16.709900: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5487] has already been set.
2018-02-15 15:33:16.709941: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5486] has already been set.
2018-02-15 15:33:16.710182: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5481] has already been set.
2018-02-15 15:33:16.710231: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5480] has already been set.
2018-02-15 15:33:16.710271: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5479] has already been set.
2018-02-15 15:33:16.710312: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5487] has already been set.
2018-02-15 15:33:16.710354: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5486] has already been set.
2018-02-15 15:33:16.710578: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5481] has already been set.
2018-02-15 15:33:16.710626: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5480] has already been set.
2018-02-15 15:33:16.710667: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5479] has already been set.
2018-02-15 15:33:16.710707: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5487] has already been set.
2018-02-15 15:33:16.710749: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5486] has already been set.
2018-02-15 15:33:16.710940: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5481] has already been set.
2018-02-15 15:33:16.710983: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5480] has already been set.
2018-02-15 15:33:16.711023: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5479] has already been set.
2018-02-15 15:33:16.711063: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5487] has already been set.
2018-02-15 15:33:16.711245: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5486] has already been set.
2018-02-15 15:33:16.711674: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5481] has already been set.
2018-02-15 15:33:16.711790: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5480] has already been set.
2018-02-15 15:33:16.711829: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5479] has already been set.
2018-02-15 15:33:16.711859: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5487] has already been set.
2018-02-15 15:33:16.711944: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5486] has already been set.
2018-02-15 15:33:16.713104: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5468] has already been set.
2018-02-15 15:33:16.720802: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5485] has already been set.
2018-02-15 15:33:16.720830: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5474] has already been set.
2018-02-15 15:33:16.720841: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5475] has already been set.
2018-02-15 15:33:16.720855: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5488] has already been set.
2018-02-15 15:33:16.720865: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5468] has already been set.
2018-02-15 15:33:16.720873: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5485] has already been set.
2018-02-15 15:33:16.720881: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5474] has already been set.
2018-02-15 15:33:16.720889: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5475] has already been set.
2018-02-15 15:33:16.720898: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5477] has already been set.
2018-02-15 15:33:16.720911: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5488] has already been set.
2018-02-15 15:33:16.720919: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5468] has already been set.
2018-02-15 15:33:16.720927: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5485] has already been set.
2018-02-15 15:33:16.720935: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5474] has already been set.
2018-02-15 15:33:16.720942: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5475] has already been set.
2018-02-15 15:33:16.720950: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5477] has already been set.
2018-02-15 15:33:16.720958: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5488] has already been set.
2018-02-15 15:33:16.720965: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5468] has already been set.
2018-02-15 15:33:16.720973: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5485] has already been set.
2018-02-15 15:33:16.721042: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5474] has already been set.
2018-02-15 15:33:16.721052: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5475] has already been set.
2018-02-15 15:33:16.721061: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5477] has already been set.
2018-02-15 15:33:16.721069: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5488] has already been set.
2018-02-15 15:33:16.721078: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5468] has already been set.
2018-02-15 15:33:16.721086: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5485] has already been set.
2018-02-15 15:33:16.721093: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5474] has already been set.
2018-02-15 15:33:16.721121: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5475] has already been set.
2018-02-15 15:33:16.721149: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5477] has already been set.
2018-02-15 15:33:16.721158: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5488] has already been set.
2018-02-15 15:33:16.721185: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5468] has already been set.
2018-02-15 15:33:16.721195: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5485] has already been set.
2018-02-15 15:33:16.721250: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5474] has already been set.
2018-02-15 15:33:16.721261: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5475] has already been set.
2018-02-15 15:33:16.721270: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5477] has already been set.
2018-02-15 15:33:16.721291: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5488] has already been set.
2018-02-15 15:33:16.721321: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5468] has already been set.
2018-02-15 15:33:16.721330: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5485] has already been set.
2018-02-15 15:33:16.721338: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5474] has already been set.
2018-02-15 15:33:16.721346: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5475] has already been set.
2018-02-15 15:33:16.721355: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5477] has already been set.
2018-02-15 15:33:16.721362: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5488] has already been set.
2018-02-15 15:33:16.721370: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5468] has already been set.
2018-02-15 15:33:16.721378: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5485] has already been set.
2018-02-15 15:33:16.721385: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5474] has already been set.
2018-02-15 15:33:16.721393: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5475] has already been set.
2018-02-15 15:33:16.721401: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5477] has already been set.
2018-02-15 15:33:16.721408: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5488] has already been set.
2018-02-15 15:33:16.721416: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5468] has already been set.
2018-02-15 15:33:16.721423: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5485] has already been set.
2018-02-15 15:33:16.721431: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5474] has already been set.
2018-02-15 15:33:16.721438: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5475] has already been set.
2018-02-15 15:33:16.721446: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5477] has already been set.
2018-02-15 15:33:16.721453: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5488] has already been set.
2018-02-15 15:33:16.721461: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5481] has already been set.
2018-02-15 15:33:16.721471: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5480] has already been set.
2018-02-15 15:33:16.721480: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5479] has already been set.
2018-02-15 15:33:16.721489: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5487] has already been set.
2018-02-15 15:33:16.721501: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[5486] has already been set.
2018-02-15 15:33:16,794 - 357 - WARNING - ERROR_TF_FALLBACK_TO_ONDEMAND_EVALUATION: Unable to resolve operation output shapes in single pass. Using on-demand evaluation!
2018-02-15 15:33:17.443825: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[0] has already been set.
2018-02-15 15:33:17.443976: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[0] has already been set.
2018-02-15 15:33:17.444030: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[0] has already been set.
2018-02-15 15:33:17.444137: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[0] has already been set.
2018-02-15 15:33:17.444188: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[0] has already been set.
2018-02-15 15:33:17.444234: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[0] has already been set.
2018-02-15 15:33:17.444280: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[0] has already been set.
2018-02-15 15:33:17.444326: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[0] has already been set.
2018-02-15 15:33:17.444372: W tensorflow/core/framework/op_kernel.cc:1198] Internal: Retval[0] has already been set.
2018-02-15 15:33:17,457 - 126 - ERROR - Encountered Error: Retval[0] has already been set.
   [[Node: _retval_Preprocessor/map/while/ResizeImage/ExpandDims_0_0 = _Retval[T=DT_FLOAT, index=0, _device="/job:localhost/replica:0/task:0/device:CPU:0"](Preprocessor/map/while/ResizeImage/ExpandDims)]]
Traceback (most recent call last):
File "/home/pragnesh/Android/snpe-1.10.1/bin/x86_64-linux-clang/snpe-tensorflow-to-dlc", line 120, in main
    converter.convert(args.dlc, args.model_version, converter_command)
File "/home/pragnesh/Android/snpe-1.10.1/lib/python/converters/tensorflow/converter.py", line 262, in convert
    self._convert_layers()
File "/home/pragnesh/Android/snpe-1.10.1/lib/python/converters/tensorflow/converter.py", line 299, in _convert_layers
    descriptors.extend(self._resolve_descriptors_from_scope(scope.name, scope.child_ops()))
File "/home/pragnesh/Android/snpe-1.10.1/lib/python/converters/tensorflow/converter.py", line 368, in _resolve_descriptors_from_scope
    candidate_descriptor = resolver.resolve_layer(scope_name, remaining_ops, self._graph_helper)
File "/home/pragnesh/Android/snpe-1.10.1/lib/python/converters/tensorflow/layers/resize.py", line 47, in resolve_layer
    input_tensor_shape = graph_helper.get_op_output_shape(input_tensor)
File "/home/pragnesh/Android/snpe-1.10.1/lib/python/converters/tensorflow/util.py", line 161, in get_op_output_shape
    shapes = self._evaluate_tensor_output_shape([tensor])
File "/home/pragnesh/Android/snpe-1.10.1/lib/python/converters/tensorflow/util.py", line 174, in _evaluate_tensor_output_shape
    outputs = self._session.run(fetches=tensors, feed_dict=input_tensors)
File "/home/pragnesh/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py", line 895, in run
    run_metadata_ptr)
File "/home/pragnesh/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py", line 1128, in _run
    feed_dict_tensor, options, run_metadata)
File "/home/pragnesh/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py", line 1344, in _do_run
    options, run_metadata)
File "/home/pragnesh/anaconda2/lib/python2.7/site-packages/tensorflow/python/client/session.py", line 1363, in _do_call
    raise type(e)(node_def, op, message)
InternalError: Retval[0] has already been set.
   [[Node: _retval_Preprocessor/map/while/ResizeImage/ExpandDims_0_0 = _Retval[T=DT_FLOAT, index=0, _device="/job:localhost/replica:0/task:0/device:CPU:0"](Preprocessor/map/while/ResizeImage/ExpandDims)]]

Is this error related to model not supported by snpe-1.10 ? if not, please help me to resolve it.

Qualcomm SNPE moderator, please let us know that, is Mobilenet_SSD is supported by SNPE 1.10? if not then when the ssd supported, SNPE release is planned?

Thanks in advance!

Forum vote up/down

Re: Errow in converting ssd_Mobilenet model tensorflow pb... #1

madhavajay

Join Date: 15 Mar 18

Posts: 22

Posted: Sun, 2018-04-29 17:21

Top

I have the 4 output tensors in Android but they dont make any sense.

Dog picture here:

https://imgur.com/a/JGDiHWo

The results for that picture of a dog look like this:

result = {ArrayList@5092} size = 50

0 = {Classifier$Recognition@5205} "[0] sports ball (4.4%) RectF(28.137482, 119.56744, 283.36014, 298.3802)"

1 = {Classifier$Recognition@5206} "[1] sports ball (4.1%) RectF(128.9198, 66.366425, 287.69986, 251.92664)"

2 = {Classifier$Recognition@5207} "[2] sports ball (4.1%) RectF(115.56224, 13.20836, 291.68512, 170.67203)"

3 = {Classifier$Recognition@5208} "[3] sports ball (3.9%) RectF(57.43878, 13.773882, 263.34735, 157.36337)"

4 = {Classifier$Recognition@5209} "[4] sports ball (3.7%) RectF(163.60475, 41.56569, 268.53174, 269.78842)"

5 = {Classifier$Recognition@5210} "[5] sports ball (3.7%) RectF(156.57309, 6.4187875, 270.69257, 207.89496)"

6 = {Classifier$Recognition@5211} "[6] sports ball (3.4%) RectF(46.86135, 148.17146, 206.738, 300.0)"

7 = {Classifier$Recognition@5212} "[7] sports ball (3.1%) RectF(39.446228, 191.30995, 215.80078, 266.1879)"

8 = {Classifier$Recognition@5213} "[8] sports ball (3.0%) RectF(89.45885, 89.07802, 297.30853, 218.85495)"

9 = {Classifier$Recognition@5214} "[9] sports ball (3.0%) RectF(116.71162, 145.27698, 280.3793, 300.0)"

10 = {Classifier$Recognition@5215} "[10] sports ball (3.0%) RectF(6.7646413, 163.43538, 300.0, 278.81375)"

Im using the same code as the TensorFlowObjectDetectionAPIModel.java from the TensorFlow SSD Mobilev1 Android Example.

private sNPEObjectDetectionAPIModel() {}

@Override
public List<Recognition> recognizeImage(final Bitmap bitmap) {
  // passing in dog.png as bitmap
  Log.v(TAG, "recognizeImage");
  // Log this method so that it can be analyzed with systrace.
  Trace.beginSection("recognizeImage");

  Trace.beginSection("preprocessBitmap");
  // Preprocess the image data from 0-255 int to normalized float based
  // on the provided parameters.
  bitmap.getPixels(intValues, 0, bitmap.getWidth(), 0, 0, bitmap.getWidth(), bitmap.getHeight());

  for (int i = 0; i < intValues.length; ++i) {
    byteValues[i * 3 + 2] = (byte) (intValues[i] & 0xFF);
    byteValues[i * 3 + 1] = (byte) ((intValues[i] >> 8) & 0xFF);
    byteValues[i * 3 + 0] = (byte) ((intValues[i] >> 16) & 0xFF);
  }
  Trace.endSection(); // preprocessBitmap

  // Run the inference call.
  Trace.beginSection("run");
  final Map<String, FloatTensor> map = new HashMap<>();
  //String inputTensorName = "image_tensor:0";
  //map.put(inputTensorName, inputTensor);
  // inputTensorName = "Preprocessor/sub:0";
  map.put(inputName, inputTensor);
  Log.v(TAG, "Got input");
  Log.v(TAG, map.toString());
  final Map<String, FloatTensor> output = inferenceInterface.execute(map);
  Trace.endSection();
  // detection_classes:0
  Log.v(TAG, "Got output");
  Log.v(TAG, output.toString());

  for (String outputName : outputNames) {
    Log.v(TAG, "for output name: " + outputName);
    output.get(outputName).read(this.outputs, 0, this.outputs.length);
  }

  // Get output tensor and write to float arrays
  output.get(SNPE_TENSOR_BOXES).read(this.outputLocations, 0, this.outputLocations.length);
  output.get(SNPE_TENSOR_SCORES).read(this.outputScores, 0, this.outputScores.length);
  output.get(SNPE_TENSOR_CLASSES).read(this.outputClasses, 0, this.outputClasses.length);
  output.get(SNPE_TENSOR_CLASSES_DETECT).read(this.outputDetectionClasses, 0, this.outputDetectionClasses.length);

The instructions of the 1.13 sNPE sdk are this:

snpe-tensorflow-to-dlc --graph ssd_mobilenet_v1_coco_2017_11_17/frozen_inference_graph.pb -i Preprocessor/sub 300,300,3 --out_node detection_classes --out_node detection_boxes --out_node detection_scores --dlc mobilenet_ssd_test.dlc --allow_unconsumed_nodes

Which is not the same as the Android example which feeds into the first operation in the original graph called: image_tensor

Converting using the tool pointing at image_tensor as the input fails.

Does this mean that there are some preprocessing steps that the normal graph in the Android SSD example is doing to the image which the dlc file does not have? If so how do I figure out what they are. Looking at the TF ops for the SSD there are a tonne of things before Preprocessor/sub:0 appears.

I added the add_6 output tensor but it didn't make any difference.

I don't think its a NHWC vs NCHW problem since SSD mobilenetv1 is NHWC and I believe the sNPE dlc format is NHWC too.
What else is going on? I have tried swapping RBG for BGR in the order of colours and that has no affect which it shouldnt really since the detection is so far off it cant get a dog.

Can anyone shed any light here?

Why are the detection % really low and for nearly every single region of the image?

In the TF example with the exact same image loaded as a bitmap and fed directly in, it only detects the dog:

[0] dog (97.7%) RectF(69.8032, 21.949268, 274.2805, 278.44122)

Can any of the people in this thread explain whats wrong or share some working code?

Thanks!

Re: Errow in converting ssd_Mobilenet model tensorflow pb... #2

stewartc

Join Date: 18 Apr 13

Posts: 1

Posted: Sun, 2018-04-29 18:36

Top

You should use a preprocessor option to specific the image dimensions:

As specified in https://developer.qualcomm.com/docs/snpe/convert_mobilenetssd.html.

Re: Errow in converting ssd_Mobilenet model tensorflow pb... #3

demo-name

Join Date: 26 Jan 18

Posts: 8

Posted: Mon, 2018-04-30 06:29

Top

@madhavajay, for their conversion tool you do need to feed the input to preprocessor/sub, which means you need to take into account the image mean and std, see the comment here: https://developer.qualcomm.com/forum/qdn-forums/software/snapdragon-neur...

Also I'm assuming you already did the crop and resize on your bitmap so it is the correct dimensions for input (300x300x3)

Good luck!

Re: Errow in converting ssd_Mobilenet model tensorflow pb... #4

madhavajay

Join Date: 15 Mar 18

Posts: 22

Posted: Mon, 2018-04-30 13:46

Top

Thanks for the help, I got it working.

Quick question:

Should I expect to see performance improvement on an LG G6 with GPU mode?

The DSP mode won't work for some reason, I have checked and I have the necessary vendor libs in my LG's folders.
But both GPU and GPU_FLOAT16 run indentically in performance to the normal TensorFlow PB SSD demo.

Surely it should be better with sNPE and GPU?

I have profiled the app with the Trepn Profiler and you can see that the GPU and CPU are both running under more load with the app running. Neither is at max though.

See trepn profiler here:

https://imgur.com/a/5ouAL0x

Seems like either the SSD with TF on Android is using GPU accel so the GPU accel in the Qualcomm sNPE DLC model isnt much different. OR... theyre both using CPU mostly which is why the performance is about 500ms per inference. Also i have seen 500ms per inference with no Trepn profiling or screen recording so thats not the cause of it.

Iv seen people here claim much better performance on DSP chips can anyone weigh in?

Re: Errow in converting ssd_Mobilenet model tensorflow pb... #5

demo-name

Join Date: 26 Jan 18

Posts: 8

Posted: Mon, 2018-04-30 14:48

Top

Glad to hear it's working. I'm not sure where, but on another thread it was mentioned that CPU fallback must be enabled because not all layers are supported by the GPU yet. So that may be what's going on. I haven't ran a profiler on my app, but I was able to get about 10 FPS detections on a Galaxy S8.

And yes, the DSP doesn't support Mobilenet+SSD yet... I'd really like to see the performance gains with that too!

I've been using SNPE 1.13.0

Cheers!

Re: Errow in converting ssd_Mobilenet model tensorflow pb... #6

madhavajay

Join Date: 15 Mar 18

Posts: 22

Posted: Mon, 2018-04-30 15:09

Top

User xavier12358 on this thread said:

https://developer.qualcomm.com/forum/qdn-forums/software/snapdragon-neur...

Hello madhavajay,

On my Snapdragon 820 64 bit:

On tensorflow time to process images is 400ms up to 600ms

On CPU with SNPE time to process images is 1100 ms

On CPU/GPU with SNPE time to process images is approximatly 77ms

On DSP, the SSD model is not compatible.

I hopes this answer your questions.

I hope thats not a typo, but if it is then what the hell is the point of having DLC conversion on a model if the performance isnt noticeably better?

The whole reason I got an 820 device (821 LG G6) and went down this road was because I assumed that the performance he mentioned of 77ms is possible which would be great!

Does anyone else have any comments on this?

This is really frustrating.

Dear Qualcomm and its staff, do you really feel this is an adequate development experience for your developers?

If it was me I would say Qualcomm should have a full Model Zoo with listed support and performance metrics on a bunch of known hardware as well as working demos including some in the app store, better docs that are actually up to date and run on more systems than Ubuntu 14, and python2 and someone who answers questions on the forum rather than let them sit there forever.

Just sayin...

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.

Sort By

Filter Results