edgeai example can't run latest pre compiled sample models

Weizhe Liu

Prodigy 30 points

Other Parts Discussed in Thread: SK-TDA4VM

Hi,

I am trying to run tidl demos on SK-TDA4VM board.

uname -a
Linux j7-evm 5.10.41-g4c2eade9f7 #1 SMP PREEMPT Thu Sep 2 18:18:01 UTC 2021 aarch64 aarch64 aarch64 GNU/Linux

edgeai-tidl-tools is using latest master branch

I am following the instructions on the bottom of https://github.com/TexasInstruments/edgeai-tidl-tools/blob/master/examples/osrt_cpp/README.md

Models are downloaded from https://github.com/TexasInstruments/edgeai-modelzoo/tree/master/modelartifacts/8bits

But I found only a few models is OK to run, most of the pre compiled models failed to init.

For example,

od-8020_onnxrt_edgeai-mmdet_ssd-lite_mobilenetv2_512x512_20201214_model_onnx is OK, but

od-8020_onnxrt_edgeai-mmdet_ssd_mobilenetv2_lite_512x512_20201214_model_onnx failed when init.

(See attachment for full log: log_good.txt and log_bad.txt)

What's the difference of above two models?

And where can I found corresponding pth file of the first model? The link is broken: https://github.com/TexasInstruments/edgeai-modelzoo/blob/release/models/vision/detection/coco/edgeai-mmdet/ssd-lite_mobilenetv2_512x512_20201214_checkpoint.pth.link

Attach1: model test results.

=========== pass ========

cl-0000_tflitert_mlperf_mobilenet_v1_1.0_224_tflite
od-8020_onnxrt_edgeai-mmdet_ssd-lite_mobilenetv2_512x512_20201214_model_onnx
od-8030_onnxrt_edgeai-mmdet_ssd-lite_mobilenetv2_fpn_512x512_20201110_model_onnx
od-8040_onnxrt_edgeai-mmdet_ssd-lite_regnetx-200mf_fpn_bgr_320x320_20201010_model_onnx
od-8050_onnxrt_edgeai-mmdet_ssd-lite_regnetx-800mf_fpn_bgr_512x512_20200919_model_onnx
od-8060_onnxrt_edgeai-mmdet_ssd-lite_regnetx-1.6gf_bifpn168x4_bgr_768x768_20201026_model_onnx
od-8080_onnxrt_edgeai-mmdet_yolov3-lite_regnetx-1.6gf_bgr_512x512_20210202_model_onnx
od-8090_onnxrt_edgeai-mmdet_retinanet-lite_regnetx-800mf_fpn_bgr_512x512_20200908_model_onnx

============= fail ===========

cl-0010_tflitert_tf1-models_mobilenet_v2_1.0_224_tflite
cl-0050_tflitert_tf1-models_resnet50_v1_tflite
cl-0060_tflitert_tf1-models_resnet50_v2_tflite
cl-6100_onnxrt_torchvision_resnet18_onnx
cl-6101_onnxrt_torchvision_resnet18_512x512_onnx
kd-7000_onnxrt_edgeai-mmpose_mobilenetv2_fpn_spp_udp_512_20210610_onnx
kd-7010_onnxrt_edgeai-mmpose_resnet50_fpn_spp_udp_512_20210610_onnx
kd-7020_onnxrt_edgeai-mmpose_mobilenetv2_pan_spp_udp_512_20210617_onnx
kd-7030_onnxrt_edgeai-mmpose_resnet50_pan_spp_udp_512_20210616_onnx
od-2000_tflitert_mlperf_ssd_mobilenet_v1_coco_20180128_tflite
od-2010_tflitert_mlperf_ssd_mobilenet_v2_300_float_tflite
od-2020_tflitert_tf1-models_ssdlite_mobiledet_dsp_320x320_coco_20200519_tflite
od-2030_tflitert_tf1-models_ssdlite_mobiledet_edgetpu_320x320_coco_20200519_tflite
od-2060_tflitert_tf1-models_ssdlite_mobilenet_v2_coco_20180509_tflite
od-2070_tflitert_tf2-models_ssd_mobilenet_v1_fpn_640x640_coco17_tpu-8_tflite
od-2080_tflitert_tf2-models_ssd_mobilenet_v2_fpnlite_320x320_coco17_tpu-8_tflite
od-2090_tflitert_tf2-models_ssd_mobilenet_v2_fpnlite_640x640_coco17_tpu-8_tflite
od-2100_tflitert_tf2-models_ssd_resnet50_v1_fpn_640x640_coco17_tpu-8_tflite
od-2110_tflitert_google-automl_efficientdet-lite0_bifpn_maxpool2x2_relu_ti-lite_tflite
od-2130_tflitert_tf2-models_ssd_mobilenet_v2_320x320_coco17_tpu-8_tflite
od-2150_tflitert_google-automl_efficientdet_lite1_relu_tflite
od-2170_tflitert_google-automl_efficientdet_lite3_relu_tflite
od-8020_onnxrt_edgeai-mmdet_ssd_mobilenetv2_lite_512x512_20201214_model_onnx
od-8030_onnxrt_edgeai-mmdet_ssd_mobilenetv2_fpn_lite_512x512_20201110_model_onnx
od-8040_onnxrt_edgeai-mmdet_ssd_regnetx-200mf_fpn_bgr_lite_320x320_20201010_model_onnx
od-8050_onnxrt_edgeai-mmdet_ssd_regnetx-800mf_fpn_bgr_lite_512x512_20200919_model_onnx
od-8060_onnxrt_edgeai-mmdet_ssd_regnetx-1.6gf_fpn_bgr_lite_768x768_20200923_model_onnx
od-8080_onnxrt_edgeai-mmdet_yolov3_regnetx-1.6gf_bgr_lite_512x512_20210202_model_onnx
od-8090_onnxrt_edgeai-mmdet_retinanet_regnetx-800mf_fpn_bgr_lite_512x512_20200908_model_onnx
od-8100_onnxrt_weights_yolov5s6_640_ti_lite_37p4_56p0_onnx
od-8110_onnxrt_weights_yolov5s6_384_ti_lite_32p8_51p2_onnx

Fullscreen log_bad.txt Download

***** Display run Config: start *****
verbose level set to: 1
accelerated mode set to: 1
device mem set to: 1
accelerated mode set to: 1
loop count set to: 1
input mean set to: 0
input std set to: 0
artifacts path set to: 
model path set to: 
image path set to: test_data/airshow.jpg
device_type set to: cpu
labels path set to: test_data/labels.txt
model zoo path set to: model_prebuild/od-8020_onnxrt_edgeai-mmdet_ssd_mobilenetv2_lite_512x512_20201214_model_onnx/
num of threads set to: 4
num of results set to: 5
num of warmup runs set to: 2
task type set to: 

***** Display run Config: end *****
[03:25:34.000.000000]:INFO:[dumpInfo:0072] InfererConfig::Model Path        = model_prebuild/od-8020_onnxrt_edgeai-mmdet_ssd_mobilenetv2_lite_512x512_20201214_model_onnx//model/ssd_mobilenetv2_lite_512x512_20201214_model.onnx
[03:25:34.000.000028]:INFO:[dumpInfo:0073] InfererConfig::Artifacts Path    = model_prebuild/od-8020_onnxrt_edgeai-mmdet_ssd_mobilenetv2_lite_512x512_20201214_model_onnx/artifacts
[03:25:34.000.000035]:INFO:[dumpInfo:0074] InfererConfig::Runtime API       = onnxrt
[03:25:34.000.000041]:INFO:[dumpInfo:0075] InfererConfig::Device Type       = CPU


[03:25:34.000.000048]:INFO:[dumpInfo:0247] PreprocessImageConfig::modelName       = od-8020_onnxrt_edgeai-mmdet_ssd_mobilenetv2_lite_512x512_20201214_model_onnx
[03:25:34.000.000053]:INFO:[dumpInfo:0248] PreprocessImageConfig::rtType          = onnxrt
[03:25:34.000.000058]:INFO:[dumpInfo:0249] PreprocessImageConfig::taskType        = detection
[03:25:34.000.000064]:INFO:[dumpInfo:0250] PreprocessImageConfig::dataLayout      = NCHW
[03:25:34.000.000069]:INFO:[dumpInfo:0251] PreprocessImageConfig::inDataWidth     = 320
[03:25:34.000.000074]:INFO:[dumpInfo:0252] PreprocessImageConfig::inDataHeight    = 240
[03:25:34.000.000080]:INFO:[dumpInfo:0253] PreprocessImageConfig::resizeWidth     = 512
[03:25:34.000.000085]:INFO:[dumpInfo:0254] PreprocessImageConfig::resizeHeight    = 512
[03:25:34.000.000091]:INFO:[dumpInfo:0255] PreprocessImageConfig::outDataWidth    = 512
[03:25:34.000.000096]:INFO:[dumpInfo:0256] PreprocessImageConfig::outDataHeight   = 512
[03:25:34.000.000101]:INFO:[dumpInfo:0257] PreprocessImageConfig::numChannels     = 3
[03:25:34.000.000107]:INFO:[dumpInfo:0259] PreprocessImageConfig::mean          = [ 123.675003 116.279999 103.529999 ]
[03:25:34.000.000124]:INFO:[dumpInfo:0267] PreprocessImageConfig::scale         = [ 0.017125 0.017507 0.017429 ]


[03:25:34.000.000135]:INFO:[dumpInfo:0369] PostprocessImageConfig::modelName      = od-8020_onnxrt_edgeai-mmdet_ssd_mobilenetv2_lite_512x512_20201214_model_onnx
[03:25:34.000.000141]:INFO:[dumpInfo:0370] PostprocessImageConfig::rtType         = onnxrt
[03:25:34.000.000146]:INFO:[dumpInfo:0371] PostprocessImageConfig::taskType       = detection
[03:25:34.000.000152]:INFO:[dumpInfo:0372] PostprocessImageConfig::dataLayout     = NCHW
[03:25:34.000.000157]:INFO:[dumpInfo:0373] PostprocessImageConfig::inDataWidth    = 512
[03:25:34.000.000162]:INFO:[dumpInfo:0374] PostprocessImageConfig::inDataHeight   = 512
[03:25:34.000.000168]:INFO:[dumpInfo:0375] PostprocessImageConfig::outDataWidth   = 1280
[03:25:34.000.000173]:INFO:[dumpInfo:0376] PostprocessImageConfig::outDataHeight  = 720
[03:25:34.000.000178]:INFO:[dumpInfo:0377] PostprocessImageConfig::dispWidth      = 1920
[03:25:34.000.000184]:INFO:[dumpInfo:0378] PostprocessImageConfig::dispHeight     = 1080
[03:25:34.000.000190]:INFO:[dumpInfo:0379] PostprocessImageConfig::vizThreshold   = 0.300000
[03:25:34.000.000197]:INFO:[dumpInfo:0380] PostprocessImageConfig::alpha          = 0.500000
[03:25:34.000.000203]:INFO:[dumpInfo:0381] PostprocessImageConfig::normDetect     = 0
[03:25:34.000.000208]:INFO:[dumpInfo:0382] PostprocessImageConfig::labelOffsetMap = [ (0, 1) (1, 2) (2, 3) (3, 4) (4, 5) (5, 6) (6, 7) (7, 8) (8, 9) (9, 10) (10, 11) (11, 13) (12, 14) (13, 15) (14, 16) (15, 17) (16, 18) (17, 19) (18, 20) (19, 21) (20, 22) (21, 23) (22, 24) (23, 25) (24, 27) (25, 28) (26, 31) (27, 32) (28, 33) (29, 34) (30, 35) (31, 36) (32, 37) (33, 38) (34, 39) (35, 40) (36, 41) (37, 42) (38, 43) (39, 44) (40, 46) (41, 47) (42, 48) (43, 49) (44, 50) (45, 51) (46, 52) (47, 53) (48, 54) (49, 55) (50, 56) (51, 57) (52, 58) (53, 59) (54, 60) (55, 61) (56, 62) (57, 63) (58, 64) (59, 65) (60, 67) (61, 70) (62, 72) (63, 73) (64, 74) (65, 75) (66, 76) (67, 77) (68, 78) (69, 79) (70, 80) (71, 81) (72, 82) (73, 84) (74, 85) (75, 86) (76, 87) (77, 88) (78, 89) (79, 90) (80, 91) ]
[03:25:34.000.000343]:INFO:[dumpInfo:0392] PostprocessImageConfig::formatter = [  0 1 2 3 5 4]
[03:25:34.000.000355]:INFO:[dumpInfo:0401] PostprocessImageConfig::resultIndices = [  0 1 2]

[03:25:34.015.015891]:INFO:[runInference:0277] accelerated mode
[03:25:34.015.015939]:INFO:[runInference:0279] artifacts: model_prebuild/od-8020_onnxrt_edgeai-mmdet_ssd_mobilenetv2_lite_512x512_20201214_model_onnx/artifacts
libtidl_onnxrt_EP loaded 0xdb55370 
Final number of subgraphs created are : 1, - Offloaded Nodes - 482, Total Nodes - 482 
Calling appInit() in TIDL-RT!
APP: Init ... !!!
MEM: Init ... !!!
MEM: Initialized DMA HEAP (fd=4) !!!
MEM: Init ... Done !!!
IPC: Init ... !!!
IPC: Init ... Done !!!
REMOTE_SERVICE: Init ... !!!
REMOTE_SERVICE: Init ... Done !!!
  1613.150664 s: GTC Frequency = 200 MHz
APP: Init ... Done !!!
  1613.150699 s:  VX_ZONE_INIT:Enabled
  1613.150702 s:  VX_ZONE_ERROR:Enabled
  1613.150704 s:  VX_ZONE_WARNING:Enabled
  1613.151254 s:  VX_ZONE_INIT:[tivxInitLocal:111] Initialization Done !!!
  1613.151456 s:  VX_ZONE_INIT:[tivxHostInitLocal:86] Initialization Done for HOST !!!
  1613.164528 s:  VX_ZONE_ERROR:[ownContextSendCmd:815] Command ack message returned failure cmd_status: -1
  1613.164535 s:  VX_ZONE_ERROR:[ownContextSendCmd:851] tivxEventWait() failed.
  1613.164540 s:  VX_ZONE_ERROR:[ownNodeKernelInit:538] Target kernel, TIVX_CMD_NODE_CREATE failed for node TIDLNode
  1613.164543 s:  VX_ZONE_ERROR:[ownNodeKernelInit:539] Please be sure the target callbacks have been registered for this core
  1613.164546 s:  VX_ZONE_ERROR:[ownNodeKernelInit:540] If the target callbacks have been registered, please ensure no errors are occurring within the create callback of this kernel
  1613.164550 s:  VX_ZONE_ERROR:[ownGraphNodeKernelInit:583] kernel init for node 0, kernel com.ti.tidl ... failed !!!
  1613.164561 s:  VX_ZONE_ERROR:[vxVerifyGraph:2044] Node kernel init failed
  1613.164564 s:  VX_ZONE_ERROR:[vxVerifyGraph:2098] Graph verify failed
TIDL_RT_OVX: ERROR: Verifying TIDL graph ... Failed !!!
TIDL_RT_OVX: ERROR: Verify OpenVX graph failed
[03:25:46.11965.965770]:INFO:[runInference:0299] Loaded model model_prebuild/od-8020_onnxrt_edgeai-mmdet_ssd_mobilenetv2_lite_512x512_20201214_model_onnx//model/ssd_mobilenetv2_lite_512x512_20201214_model.onnx
[03:25:46.11965.965869]:INFO:[printTensorInfo:0211] number of inputs:1 
[03:25:46.11965.965879]:INFO:[printTensorInfo:0212] number of outputs: [03:25:46.11965.965886]:INFO:[printTensorInfo:0213] input(0) name: [03:25:46.11965.965893]:INFO:[printTensorInfo:0218] Input 0 : name=P͠
[03:25:46.11965.965901]:INFO:[printTensorInfo:0225] Input 0 : type=1
[03:25:46.11965.965910]:INFO:[printTensorInfo:0228] Input 0 : num_dims=4
[03:25:46.11965.965917]:INFO:[printTensorInfo:0231] Input 0 : dim 0=1
[03:25:46.11965.965924]:INFO:[printTensorInfo:0231] Input 0 : dim 1=3
[03:25:46.11965.965931]:INFO:[printTensorInfo:0231] Input 0 : dim 2=512
[03:25:46.11965.965937]:INFO:[printTensorInfo:0231] Input 0 : dim 3=512
[03:25:46.11968.968608]:INFO:[preprocImage:0127] template NCHW
[03:25:46.11969.969464]:INFO:[runInference:0365] Output name -- boxes 
[03:25:46.11969.969485]:INFO:[runInference:0368] create cpu done
[03:25:46.11969.969570]:INFO:[runInference:0419] Session.Run() - Started for warmup runs
  1624.736323 s:  VX_ZONE_ERROR:[ownContextSendCmd:815] Command ack message returned failure cmd_status: -1
  1624.736330 s:  VX_ZONE_ERROR:[ownContextSendCmd:851] tivxEventWait() failed.
  1624.736335 s:  VX_ZONE_ERROR:[ownNodeKernelInit:538] Target kernel, TIVX_CMD_NODE_CREATE failed for node TIDLNode
  1624.736337 s:  VX_ZONE_ERROR:[ownNodeKernelInit:539] Please be sure the target callbacks have been registered for this core
  1624.736407 s:  VX_ZONE_ERROR:[ownNodeKernelInit:540] If the target callbacks have been registered, please ensure no errors are occurring within the create callback of this kernel
  1624.736411 s:  VX_ZONE_ERROR:[ownGraphNodeKernelInit:583] kernel init for node 0, kernel com.ti.tidl ... failed !!!
  1624.736417 s:  VX_ZONE_ERROR:[vxVerifyGraph:2044] Node kernel init failed
  1624.736419 s:  VX_ZONE_ERROR:[vxVerifyGraph:2098] Graph verify failed
  1624.736543 s:  VX_ZONE_ERROR:[ownGraphScheduleGraphWrapper:820] graph is not in a state required to be scheduled
  1624.736546 s:  VX_ZONE_ERROR:[vxProcessGraph:755] schedule graph failed
  1624.736548 s:  VX_ZONE_ERROR:[vxProcessGraph:760] wait graph failed
  1624.744217 s:  VX_ZONE_ERROR:[ownContextSendCmd:815] Command ack message returned failure cmd_status: -1
  1624.744222 s:  VX_ZONE_ERROR:[ownContextSendCmd:851] tivxEventWait() failed.
  1624.744225 s:  VX_ZONE_ERROR:[ownNodeKernelInit:538] Target kernel, TIVX_CMD_NODE_CREATE failed for node TIDLNode
  1624.744227 s:  VX_ZONE_ERROR:[ownNodeKernelInit:539] Please be sure the target callbacks have been registered for this core
  1624.744230 s:  VX_ZONE_ERROR:[ownNodeKernelInit:540] If the target callbacks have been registered, please ensure no errors are occurring within the create callback of this kernel
  1624.744234 s:  VX_ZONE_ERROR:[ownGraphNodeKernelInit:583] kernel init for node 0, kernel com.ti.tidl ... failed !!!
  1624.744238 s:  VX_ZONE_ERROR:[vxVerifyGraph:2044] Node kernel init failed
  1624.744240 s:  VX_ZONE_ERROR:[vxVerifyGraph:2098] Graph verify failed
  1624.744361 s:  VX_ZONE_ERROR:[ownGraphScheduleGraphWrapper:820] graph is not in a state required to be scheduled
  1624.744364 s:  VX_ZONE_ERROR:[vxProcessGraph:755] schedule graph failed
  1624.744366 s:  VX_ZONE_ERROR:[vxProcessGraph:760] wait graph failed
  1624.751753 s:  VX_ZONE_ERROR:[ownContextSendCmd:815] Command ack message returned failure cmd_status: -1
  1624.751758 s:  VX_ZONE_ERROR:[ownContextSendCmd:851] tivxEventWait() failed.
  1624.751762 s:  VX_ZONE_ERROR:[ownNodeKernelInit:538] Target kernel, TIVX_CMD_NODE_CREATE failed for node TIDLNode
  1624.751764 s:  VX_ZONE_ERROR:[ownNodeKernelInit:539] Please be sure the target callbacks have been registered for this core
  1624.751767 s:  VX_ZONE_ERROR:[ownNodeKernelInit:540] If the target callbacks have been registered, please ensure no errors are occurring within the create callback of this kernel
  1624.751770 s:  VX_ZONE_ERROR:[ownGraphNodeKernelInit:583] kernel init for node 0, kernel com.ti.tidl ... failed !!!
  1624.751774 s:  VX_ZONE_ERROR:[vxVerifyGraph:2044] Node kernel init failed
  1624.751776 s:  VX_ZONE_ERROR:[vxVerifyGraph:2098] Graph verify failed
  1624.751896 s:  VX_ZONE_ERROR:[ownGraphScheduleGraphWrapper:820] graph is not in a state required to be scheduled
  1624.751899 s:  VX_ZONE_ERROR:[vxProcessGraph:755] schedule graph failed
  1624.751901 s:  VX_ZONE_ERROR:[vxProcessGraph:760] wait graph failed
[03:25:46.11993.993594]:INFO:[runInference:0435] invoked
[03:25:46.11993.993632]:INFO:[runInference:0436] average time: 7.523000 ms 
[03:25:46.11993.993844]:INFO:[prepDetectionResult:0439] preparing detection result 
[03:25:46.12000.000868]:INFO:[runInference:0540] Done!
Calling appDeInit() in TIDL-RT!
  1624.759820 s:  VX_ZONE_INIT:[tivxHostDeInitLocal:100] De-Initialization Done for HOST !!!
  1624.764170 s:  VX_ZONE_INIT:[tivxDeInitLocal:155] De-Initialization Done !!!
APP: Deinit ... !!!
REMOTE_SERVICE: Deinit ... !!!
REMOTE_SERVICE: Deinit ... Done !!!
IPC: Deinit ... !!!
IPC: DeInit ... Done !!!
MEM: Deinit ... !!!
MEM: Alloc's: 10 alloc's of 10939592 bytes 
MEM: Free's : 10 free's  of 10939592 bytes 
MEM: Open's : 0 allocs  of 0 bytes 
MEM: Deinit ... Done !!!
APP: Deinit ... Done !!!

Fullscreen log_good.txt Download

***** Display run Config: start *****
verbose level set to: 1
accelerated mode set to: 1
device mem set to: 1
accelerated mode set to: 1
loop count set to: 1
input mean set to: 0
input std set to: 0
artifacts path set to: 
model path set to: 
image path set to: test_data/airshow.jpg
device_type set to: cpu
labels path set to: test_data/labels.txt
model zoo path set to: model_prebuild/od-8020_onnxrt_edgeai-mmdet_ssd-lite_mobilenetv2_512x512_20201214_model_onnx/
num of threads set to: 4
num of results set to: 5
num of warmup runs set to: 2
task type set to: 

***** Display run Config: end *****
[03:26:08.000.000000]:INFO:[dumpInfo:0072] InfererConfig::Model Path        = model_prebuild/od-8020_onnxrt_edgeai-mmdet_ssd-lite_mobilenetv2_512x512_20201214_model_onnx//model/ssd-lite_mobilenetv2_512x512_20201214_model.onnx
[03:26:08.000.000031]:INFO:[dumpInfo:0073] InfererConfig::Artifacts Path    = model_prebuild/od-8020_onnxrt_edgeai-mmdet_ssd-lite_mobilenetv2_512x512_20201214_model_onnx/artifacts
[03:26:08.000.000038]:INFO:[dumpInfo:0074] InfererConfig::Runtime API       = onnxrt
[03:26:08.000.000044]:INFO:[dumpInfo:0075] InfererConfig::Device Type       = CPU


[03:26:08.000.000051]:INFO:[dumpInfo:0247] PreprocessImageConfig::modelName       = od-8020_onnxrt_edgeai-mmdet_ssd-lite_mobilenetv2_512x512_20201214_model_onnx
[03:26:08.000.000057]:INFO:[dumpInfo:0248] PreprocessImageConfig::rtType          = onnxrt
[03:26:08.000.000063]:INFO:[dumpInfo:0249] PreprocessImageConfig::taskType        = detection
[03:26:08.000.000068]:INFO:[dumpInfo:0250] PreprocessImageConfig::dataLayout      = NCHW
[03:26:08.000.000074]:INFO:[dumpInfo:0251] PreprocessImageConfig::inDataWidth     = 320
[03:26:08.000.000080]:INFO:[dumpInfo:0252] PreprocessImageConfig::inDataHeight    = 240
[03:26:08.000.000086]:INFO:[dumpInfo:0253] PreprocessImageConfig::resizeWidth     = 512
[03:26:08.000.000091]:INFO:[dumpInfo:0254] PreprocessImageConfig::resizeHeight    = 512
[03:26:08.000.000097]:INFO:[dumpInfo:0255] PreprocessImageConfig::outDataWidth    = 512
[03:26:08.000.000103]:INFO:[dumpInfo:0256] PreprocessImageConfig::outDataHeight   = 512
[03:26:08.000.000108]:INFO:[dumpInfo:0257] PreprocessImageConfig::numChannels     = 3
[03:26:08.000.000114]:INFO:[dumpInfo:0259] PreprocessImageConfig::mean          = [ 123.675003 116.279999 103.529999 ]
[03:26:08.000.000130]:INFO:[dumpInfo:0267] PreprocessImageConfig::scale         = [ 0.017000 0.018000 0.017000 ]


[03:26:08.000.000141]:INFO:[dumpInfo:0369] PostprocessImageConfig::modelName      = od-8020_onnxrt_edgeai-mmdet_ssd-lite_mobilenetv2_512x512_20201214_model_onnx
[03:26:08.000.000147]:INFO:[dumpInfo:0370] PostprocessImageConfig::rtType         = onnxrt
[03:26:08.000.000153]:INFO:[dumpInfo:0371] PostprocessImageConfig::taskType       = detection
[03:26:08.000.000159]:INFO:[dumpInfo:0372] PostprocessImageConfig::dataLayout     = NCHW
[03:26:08.000.000164]:INFO:[dumpInfo:0373] PostprocessImageConfig::inDataWidth    = 512
[03:26:08.000.000170]:INFO:[dumpInfo:0374] PostprocessImageConfig::inDataHeight   = 512
[03:26:08.000.000176]:INFO:[dumpInfo:0375] PostprocessImageConfig::outDataWidth   = 1280
[03:26:08.000.000182]:INFO:[dumpInfo:0376] PostprocessImageConfig::outDataHeight  = 720
[03:26:08.000.000187]:INFO:[dumpInfo:0377] PostprocessImageConfig::dispWidth      = 1920
[03:26:08.000.000193]:INFO:[dumpInfo:0378] PostprocessImageConfig::dispHeight     = 1080
[03:26:08.000.000199]:INFO:[dumpInfo:0379] PostprocessImageConfig::vizThreshold   = 0.300000
[03:26:08.000.000234]:INFO:[dumpInfo:0380] PostprocessImageConfig::alpha          = 0.500000
[03:26:08.000.000242]:INFO:[dumpInfo:0381] PostprocessImageConfig::normDetect     = 0
[03:26:08.000.000248]:INFO:[dumpInfo:0382] PostprocessImageConfig::labelOffsetMap = [ (0, 1) (1, 2) (2, 3) (3, 4) (4, 5) (5, 6) (6, 7) (7, 8) (8, 9) (9, 10) (10, 11) (11, 13) (12, 14) (13, 15) (14, 16) (15, 17) (16, 18) (17, 19) (18, 20) (19, 21) (20, 22) (21, 23) (22, 24) (23, 25) (24, 27) (25, 28) (26, 31) (27, 32) (28, 33) (29, 34) (30, 35) (31, 36) (32, 37) (33, 38) (34, 39) (35, 40) (36, 41) (37, 42) (38, 43) (39, 44) (40, 46) (41, 47) (42, 48) (43, 49) (44, 50) (45, 51) (46, 52) (47, 53) (48, 54) (49, 55) (50, 56) (51, 57) (52, 58) (53, 59) (54, 60) (55, 61) (56, 62) (57, 63) (58, 64) (59, 65) (60, 67) (61, 70) (62, 72) (63, 73) (64, 74) (65, 75) (66, 76) (67, 77) (68, 78) (69, 79) (70, 80) (71, 81) (72, 82) (73, 84) (74, 85) (75, 86) (76, 87) (77, 88) (78, 89) (79, 90) (80, 91) ]
[03:26:08.000.000378]:INFO:[dumpInfo:0392] PostprocessImageConfig::formatter = [  0 1 2 3 5 4]
[03:26:08.000.000390]:INFO:[dumpInfo:0401] PostprocessImageConfig::resultIndices = [  0 1 2]

[03:26:08.016.016921]:INFO:[runInference:0277] accelerated mode
[03:26:08.016.016971]:INFO:[runInference:0279] artifacts: model_prebuild/od-8020_onnxrt_edgeai-mmdet_ssd-lite_mobilenetv2_512x512_20201214_model_onnx/artifacts
libtidl_onnxrt_EP loaded 0x8d75550 
Final number of subgraphs created are : 1, - Offloaded Nodes - 482, Total Nodes - 482 
Calling appInit() in TIDL-RT!
APP: Init ... !!!
MEM: Init ... !!!
MEM: Initialized DMA HEAP (fd=4) !!!
MEM: Init ... Done !!!
IPC: Init ... !!!
IPC: Init ... Done !!!
REMOTE_SERVICE: Init ... !!!
REMOTE_SERVICE: Init ... Done !!!
  1647.982915 s: GTC Frequency = 200 MHz
APP: Init ... Done !!!
  1647.982968 s:  VX_ZONE_INIT:Enabled
  1647.982970 s:  VX_ZONE_ERROR:Enabled
  1647.982973 s:  VX_ZONE_WARNING:Enabled
  1647.983798 s:  VX_ZONE_INIT:[tivxInitLocal:111] Initialization Done !!!
  1647.984076 s:  VX_ZONE_INIT:[tivxHostInitLocal:86] Initialization Done for HOST !!!
[03:26:21.12402.402575]:INFO:[runInference:0299] Loaded model model_prebuild/od-8020_onnxrt_edgeai-mmdet_ssd-lite_mobilenetv2_512x512_20201214_model_onnx//model/ssd-lite_mobilenetv2_512x512_20201214_model.onnx
[03:26:21.12402.402668]:INFO:[printTensorInfo:0211] number of inputs:1 
[03:26:21.12402.402677]:INFO:[printTensorInfo:0212] number of outputs: [03:26:21.12402.402684]:INFO:[printTensorInfo:0213] input(0) name: [03:26:21.12402.402691]:INFO:[printTensorInfo:0218] Input 0 : name= �
[03:26:21.12402.402699]:INFO:[printTensorInfo:0225] Input 0 : type=1
[03:26:21.12402.402707]:INFO:[printTensorInfo:0228] Input 0 : num_dims=4
[03:26:21.12402.402715]:INFO:[printTensorInfo:0231] Input 0 : dim 0=1
[03:26:21.12402.402722]:INFO:[printTensorInfo:0231] Input 0 : dim 1=3
[03:26:21.12402.402728]:INFO:[printTensorInfo:0231] Input 0 : dim 2=512
[03:26:21.12402.402735]:INFO:[printTensorInfo:0231] Input 0 : dim 3=512
[03:26:21.12405.405449]:INFO:[preprocImage:0127] template NCHW
[03:26:21.12406.406358]:INFO:[runInference:0365] Output name -- boxes 
[03:26:21.12406.406382]:INFO:[runInference:0368] create cpu done
[03:26:21.12406.406468]:INFO:[runInference:0419] Session.Run() - Started for warmup runs
[03:26:21.12423.423277]:INFO:[runInference:0435] invoked
[03:26:21.12423.423315]:INFO:[runInference:0436] average time: 5.334000 ms 
[03:26:21.12423.423520]:INFO:[prepDetectionResult:0439] preparing detection result 
[03:26:21.12437.437353]:INFO:[runInference:0540] Done!
Calling appDeInit() in TIDL-RT!
  1659.670800 s:  VX_ZONE_INIT:[tivxHostDeInitLocal:100] De-Initialization Done for HOST !!!
  1659.675142 s:  VX_ZONE_INIT:[tivxDeInitLocal:155] De-Initialization Done !!!
APP: Deinit ... !!!
REMOTE_SERVICE: Deinit ... !!!
REMOTE_SERVICE: Deinit ... Done !!!
IPC: Deinit ... !!!
IPC: DeInit ... Done !!!
MEM: Deinit ... !!!
MEM: Alloc's: 10 alloc's of 7908449 bytes 
MEM: Free's : 10 free's  of 7908449 bytes 
MEM: Open's : 0 allocs  of 0 bytes 
MEM: Deinit ... Done !!!
APP: Deinit ... Done !!!

over 3 years ago