Demo compatibility with jetpack 4.3

[rahairi]

Look like the demo cannot be run with tensorRT 6.0.1. Tried detectnet-camera not working.

[dusty-nv]

What error do you get? I am able to run it with JetPack 4.3.

[rahairi]

no camera output, never ending loop on tactic

[gstreamer] initialized gstreamer, version 1.14.5.0 [gstreamer] gstCamera attempting to initialize with GST_SOURCE_NVARGUS, camera 0 [gstreamer] gstCamera pipeline string: nvarguscamerasrc sensor-id=0 ! video/x-raw(memory:NVMM), width=(int)1280, height=(int)720, framerate=30/1, format=(string)NV12 ! nvvidconv flip-method=2 ! video/x-raw ! appsink name=mysink [gstreamer] gstCamera successfully initialized with GST_SOURCE_NVARGUS, camera 0

detectnet-camera: successfully initialized camera device width: 1280 height: 720 depth: 12 (bpp)

detectNet -- loading detection network model from: -- model networks/SSD-Mobilenet-v2/ssd_mobilenet_v2_coco.uff -- input_blob 'Input' -- output_blob 'NMS' -- output_count 'NMS_1' -- class_labels networks/SSD-Mobilenet-v2/ssd_coco_labels.txt -- threshold 0.500000 -- batch_size 1

[TRT] TensorRT version 6.0.1 [TRT] loading NVIDIA plugins... [TRT] Plugin Creator registration succeeded - GridAnchor_TRT [TRT] Plugin Creator registration succeeded - GridAnchorRect_TRT [TRT] Plugin Creator registration succeeded - NMS_TRT [TRT] Plugin Creator registration succeeded - Reorg_TRT [TRT] Plugin Creator registration succeeded - Region_TRT [TRT] Plugin Creator registration succeeded - Clip_TRT [TRT] Plugin Creator registration succeeded - LReLU_TRT [TRT] Plugin Creator registration succeeded - PriorBox_TRT [TRT] Plugin Creator registration succeeded - Normalize_TRT [TRT] Plugin Creator registration succeeded - RPROI_TRT [TRT] Plugin Creator registration succeeded - BatchedNMS_TRT [TRT] Could not register plugin creator: FlattenConcat_TRT in namespace: [TRT] completed loading NVIDIA plugins. [TRT] detected model format - UFF (extension '.uff') [TRT] desired precision specified for GPU: FASTEST [TRT] requested fasted precision for device GPU without providing valid calibrator, disabling INT8 [TRT] native precisions detected for GPU: FP32, FP16 [TRT] selecting fastest native precision for GPU: FP16 [TRT] attempting to open engine cache file networks/SSD-Mobilenet-v2/ssd_mobilenet_v2_coco.uff.1.1.GPU.FP16.engine [TRT] cache file not found, profiling network model on device GPU [TRT] device GPU, loading /usr/local/bin/ networks/SSD-Mobilenet-v2/ssd_mobilenet_v2_coco.uff [TRT] UFFParser: Parsing Input[Op: Input]. [TRT] UFFParser: Input -> [3,300,300] [TRT] UFFParser: Applying order forwarding to: Input [TRT] UFFParser: Parsing FeatureExtractor/MobilenetV2/Conv/weights[Op: Const]. [TRT] UFFParser: FeatureExtractor/MobilenetV2/Conv/weights -> [3,3,3,32] [TRT] UFFParser: Applying order forwarding to: FeatureExtractor/MobilenetV2/Conv/weights [TRT] UFFParser: Parsing FeatureExtractor/MobilenetV2/Conv/Conv2D[Op: Conv]. Inputs: Input, FeatureExtractor/MobilenetV2/Conv/weights [TRT] UFFParser: FeatureExtractor/MobilenetV2/Conv/Conv2D -> [32,150,150] [TRT] UFFParser: Applying order forwarding to: FeatureExtractor/MobilenetV2/Conv/Conv2D [TRT] UFFParser: Parsing FeatureExtractor/MobilenetV2/Conv/BatchNorm/moving_variance[Op: Const]. [TRT] UFFParser: FeatureExtractor/MobilenetV2/Conv/BatchNorm/moving_variance -> [32] [TRT] UFFParser: Applying order forwarding to: FeatureExtractor/MobilenetV2/Conv/BatchNorm/moving_variance [TRT] UFFParser: Parsing FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/add/y[Op: Const]. [TRT] UFFParser: FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/add/y -> [] [TRT] UFFParser: Applying order forwarding to: FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/add/y [TRT] UFFParser: Parsing FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/add[Op: Binary]. Inputs: FeatureExtractor/MobilenetV2/Conv/BatchNorm/moving_variance, FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/add/y [TRT] UFFParser: Applying order forwarding to: FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/add [TRT] UFFParser: Parsing FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/Rsqrt[Op: Unary]. Inputs: FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/add [TRT] UFFParser: Applying order forwarding to: FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/Rsqrt [TRT] UFFParser: Parsing FeatureExtractor/MobilenetV2/Conv/BatchNorm/gamma[Op: Const]. [TRT] UFFParser: FeatureExtractor/MobilenetV2/Conv/BatchNorm/gamma -> [32] [TRT] UFFParser: Applying order forwarding to: FeatureExtractor/MobilenetV2/Conv/BatchNorm/gamma [TRT] UFFParser: Parsing FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/mul[Op: Binary]. Inputs: FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/Rsqrt, FeatureExtractor/MobilenetV2/Conv/BatchNorm/gamma [TRT] UFFParser: Applying order forwarding to: FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/mul [TRT] UFFParser: Parsing FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/mul_1[Op: Binary]. Inputs: FeatureExtractor/MobilenetV2/Conv/Conv2D, FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/mul [TRT] UFFParser: FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/mul_1 -> [32,150,150] [TRT] UFFParser: Applying order forwarding to: FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/mul_1 [TRT] UFFParser: Parsing FeatureExtractor/MobilenetV2/Conv/BatchNorm/beta[Op: Const]. [TRT] UFFParser: FeatureExtractor/MobilenetV2/Conv/BatchNorm/beta -> [32] [TRT] UFFParser: Applying order forwarding to: FeatureExtractor/MobilenetV2/Conv/BatchNorm/beta [TRT] UFFParser: Parsing FeatureExtractor/MobilenetV2/Conv/BatchNorm/moving_mean[Op: Const]. [TRT] UFFParser: FeatureExtractor/MobilenetV2/Conv/BatchNorm/moving_mean -> [32] [TRT] UFFParser: Applying order forwarding to: FeatureExtractor/MobilenetV2/Conv/BatchNorm/moving_mean [TRT] UFFParser: Parsing FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/mul_2[Op: Binary]. Inputs: FeatureExtractor/MobilenetV2/Conv/BatchNorm/moving_mean, FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/mul [TRT] UFFParser: Applying order forwarding to: FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/mul_2 [TRT] UFFParser: Parsing FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/sub[Op: Binary]. Inputs: FeatureExtractor/MobilenetV2/Conv/BatchNorm/beta, FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/mul_2 [TRT] UFFParser: Applying order forwarding to: FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/sub [TRT] UFFParser: Parsing FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/add_1[Op: Binary]. Inputs: FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/mul_1, FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/sub [TRT] UFFParser: FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/add_1 -> [32,150,150] [TRT] UFFParser: Applying order forwarding to: FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/add_1 [TRT] UFFParser: Parsing FeatureExtractor/MobilenetV2/Conv/Relu6[Op: Activation]. Inputs: FeatureExtractor/MobilenetV2/Conv/BatchNorm/batchnorm/add_1 [TRT] Setting dynamic range for (Unnamed Layer* 9) [Activation]_output to [0,6] [TRT] UFFParser: FeatureExtractor/MobilenetV2/Conv/Relu6 -> [32,150,150] [TRT] UFFParser: Applying order forwarding to: FeatureExtractor/MobilenetV2/Conv/Relu6 [TRT] UFFParser: Parsing FeatureExtractor/MobilenetV2/expanded_conv/depthwise/depthwise_weights[Op: Const]. [TRT] UFFParser: FeatureExtractor/MobilenetV2/expanded_conv/depthwise/depthwise_weights -> [3,3,32,1] [TRT] UFFParser: Applying order forwarding to: FeatureExtractor/MobilenetV2/expanded_conv/depthwise/depthwise_weights [TRT] UFFParser: Parsing FeatureExtractor/MobilenetV2/expanded_conv/depthwise/depthwise[Op: Conv]. Inputs: FeatureExtractor/MobilenetV2/Conv/Relu6, FeatureExtractor/MobilenetV2/expanded_conv/depthwise/depthwise_weights [TRT] UFFParser: FeatureExtractor/MobilenetV2/expanded_conv/depthwise/depthwise -> [32,150,150] [TRT] UFFParser: Applying order forwarding to: FeatureExtractor/MobilenetV2/expanded_conv/depthwise/depthwise [TRT] UFFParser: Parsing FeatureExtractor/MobilenetV2/expanded_conv/depthwise/BatchNorm/moving_variance[Op: Const]. [TRT] UFFParser: FeatureExtractor/MobilenetV2/expanded_conv/depthwise/BatchNorm/moving_variance -> [32] [TRT] UFFParser: Applying order forwarding to: FeatureExtractor/MobilenetV2/expanded_conv/depthwise/BatchNorm/moving_variance [TRT] UFFParser: Parsing FeatureExtractor/MobilenetV2/expanded_conv/depthwise/BatchNorm/batchnorm/add/y[Op: Const]. [TRT] UFFParser: FeatureExtractor/MobilenetV2/expanded_conv/depthwise/BatchNorm/batchnorm/add/y -> [] [TRT] UFFParser: Applying order forwarding to: FeatureExtractor/MobilenetV2/expanded_conv/depthwise/BatchNorm/batchnorm/add/y

[rahairi]

• FeatureExtractor/MobilenetV2/Conv/Relu6 (CudaDepthwiseConvolution) [TRT] CudaDepthwiseConvolution has no valid tactics for this config, skipping [TRT] >>>>>>>>>>>>>>> Chose Runner Type: CaskConvolution Tactic: 1062367460111450758 [TRT] FeatureExtractor/MobilenetV2/Conv/Conv2D + FeatureExtractor/MobilenetV2/Conv/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_medium_nn_v1 [TRT]
  [TRT] Autotuning format combination: Half(1,300,90000,270000) -> Half(1,150,22500,720000) [TRT] --------------- Timing Runner: FeatureExtractor/MobilenetV2/Conv/Conv2D + FeatureExtractor/MobilenetV2/Conv/Relu6 (FusedConvActConvolution) [TRT] FusedConvActConvolution has no valid tactics for this config, skipping [TRT] --------------- Timing Runner: FeatureExtractor/MobilenetV2/Conv/Conv2D + FeatureExtractor/MobilenetV2/Conv/Relu6 (CaskConvolution) [TRT] CaskConvolution has no valid tactics for this config, skipping [TRT] --------------- Timing Runner: FeatureExtractor/MobilenetV2/Conv/Conv2D + FeatureExtractor/MobilenetV2/Conv/Relu6 (CudaConvolution) [TRT] Tactic: 0 time 2.02477 [TRT] Tactic: 1 time 1.59526 [TRT] Tactic: 2 time 2.3631 [TRT] Fastest Tactic: 1 Time: 1.59526 [TRT] --------------- Timing Runner: FeatureExtractor/MobilenetV2/Conv/Conv2D + FeatureExtractor/MobilenetV2/Conv/Relu6 (CudaDepthwiseConvolution) [TRT] CudaDepthwiseConvolution has no valid tactics for this config, skipping [TRT] >>>>>>>>>>>>>>> Chose Runner Type: CudaConvolution Tactic: 1 [TRT]
  [TRT] Autotuning format combination: Half(1,300,90000:2,180000) -> Half(1,150,22500:2,360000) [TRT] --------------- Timing Runner: FeatureExtractor/MobilenetV2/Conv/Conv2D + FeatureExtractor/MobilenetV2/Conv/Relu6 (LegacySASSConvolution) [TRT] LegacySASSConvolution has no valid tactics for this config, skipping [TRT] --------------- Timing Runner: FeatureExtractor/MobilenetV2/Conv/Conv2D + FeatureExtractor/MobilenetV2/Conv/Relu6 (FusedConvActConvolution) [TRT] FusedConvActConvolution has no valid tactics for this config, skipping [TRT] --------------- Timing Runner: FeatureExtractor/MobilenetV2/Conv/Conv2D + FeatureExtractor/MobilenetV2/Conv/Relu6 (CaskConvolution) [TRT] FeatureExtractor/MobilenetV2/Conv/Conv2D + FeatureExtractor/MobilenetV2/Conv/Relu6 (hcudnn) Set Tactic Name: maxwell_fp16x2_hcudnn_fp16x2_128x32_relu_medium_nn_v1 [TRT] Tactic: 3564772625446233998 time 0.55362 [TRT] FeatureExtractor/MobilenetV2/Conv/Conv2D + FeatureExtractor/MobilenetV2/Conv/Relu6 (hcudnn) Set Tactic Name: maxwell_fp16x2_hcudnn_fp16x2_128x32_relu_large_nn_v1 [TRT] Tactic: 3650389455493082349 time 0.581198 [TRT] FeatureExtractor/MobilenetV2/Conv/Conv2D + FeatureExtractor/MobilenetV2/Conv/Relu6 (hcudnn) Set Tactic Name: maxwell_fp16x2_hcudnn_fp16x2_128x32_relu_small_nn_v1 [TRT] Tactic: 5319956359050645452 time 0.52362 [TRT] FeatureExtractor/MobilenetV2/Conv/Conv2D + FeatureExtractor/MobilenetV2/Conv/Relu6 (hcudnn) Set Tactic Name: maxwell_fp16x2_hcudnn_fp16x2_128x64_relu_medium_nn_v1 [TRT] Tactic: 7205456024582378848 time 0.810052 [TRT] FeatureExtractor/MobilenetV2/Conv/Conv2D + FeatureExtractor/MobilenetV2/Conv/Relu6 (hcudnn) Set Tactic Name: maxwell_fp16x2_hcudnn_fp16x2_128x64_relu_large_nn_v1 [TRT] Tactic: -6490690591794140522 time 0.85289 [TRT] FeatureExtractor/MobilenetV2/Conv/Conv2D + FeatureExtractor/MobilenetV2/Conv/Relu6 (hcudnn) Set Tactic Name: maxwell_fp16x2_hcudnn_fp16x2_128x128_relu_large_nn_v1 [TRT] Tactic: -4686027666808657977 time 1.62779 [TRT] FeatureExtractor/MobilenetV2/Conv/Conv2D + FeatureExtractor/MobilenetV2/Conv/Relu6 (hcudnn) Set Tactic Name: maxwell_fp16x2_hcudnn_fp16x2_128x128_relu_small_nn_v1 [TRT] Tactic: -4212163711445252890 time 1.59414 [TRT] FeatureExtractor/MobilenetV2/Conv/Conv2D + FeatureExtractor/MobilenetV2/Conv/Relu6 (hcudnn) Set Tactic Name: maxwell_fp16x2_hcudnn_fp16x2_128x128_relu_medium_nn_v1 [TRT] Tactic: -3898373634979201110 time 1.59753 [TRT] FeatureExtractor/MobilenetV2/Conv/Conv2D + FeatureExtractor/MobilenetV2/Conv/Relu6 (hcudnn) Set Tactic Name: maxwell_fp16x2_hcudnn_fp16x2_128x64_relu_small_nn_v1 [TRT] Tactic: -2409163523992614473 time 0.808854 [TRT] Fastest Tactic: 5319956359050645452 Time: 0.52362 [TRT] --------------- Timing Runner: FeatureExtractor/MobilenetV2/Conv/Conv2D + FeatureExtractor/MobilenetV2/Conv/Relu6 (CudaConvolution) [TRT] CudaConvolution has no valid tactics for this config, skipping [TRT] --------------- Timing Runner: FeatureExtractor/MobilenetV2/Conv/Conv2D + FeatureExtractor/MobilenetV2/Conv/Relu6 (CudaDepthwiseConvolution) [TRT] CudaDepthwiseConvolution has no valid tactics for this config, skipping [TRT] >>>>>>>>>>>>>>> Chose Runner Type: CaskConvolution Tactic: 5319956359050645452 [TRT] FeatureExtractor/MobilenetV2/Conv/Conv2D + FeatureExtractor/MobilenetV2/Conv/Relu6 (hcudnn) Set Tactic Name: maxwell_fp16x2_hcudnn_fp16x2_128x32_relu_small_nn_v1 [TRT]
  [TRT] --------------- Timing Runner: (Reformat) [TRT] Tactic: 1002 time 0.548594 [TRT] Tactic: 0 time 0.799531 [TRT] Fastest Tactic: 1002 Time: 0.548594 [TRT] --------------- Timing Runner: (Reformat) [TRT] Tactic: 1002 time 0.535573 [TRT] Tactic: 0 time 0.678542 [TRT] Fastest Tactic: 1002 Time: 0.535573 [TRT] --------------- Timing Runner: (Reformat) [TRT] Tactic: 1002 time 2.16971 [TRT] Tactic: 0 time 0.58461 [TRT] Fastest Tactic: 0 Time: 0.58461 [TRT] --------------- Timing Runner: (Reformat) [TRT] Tactic: 1002 time 1.99083 [TRT] Tactic: 0 time 0.526198 [TRT] Fastest Tactic: 0 Time: 0.526198 [TRT] Autotuning format combination: Float(1,150,22500,720000) -> Float(1,150,22500,720000) [TRT] --------------- Timing Runner: FeatureExtractor/MobilenetV2/expanded_conv/depthwise/depthwise + FeatureExtractor/MobilenetV2/expanded_conv/depthwise/Relu6 (LegacySASSConvolution) [TRT] LegacySASSConvolution has no valid tactics for this config, skipping [TRT] --------------- Timing Runner: FeatureExtractor/MobilenetV2/expanded_conv/depthwise/depthwise + FeatureExtractor/MobilenetV2/expanded_conv/depthwise/Relu6 (FusedConvActConvolution) [TRT] FusedConvActConvolution has no valid tactics for this config, skipping [TRT] --------------- Timing Runner: FeatureExtractor/MobilenetV2/expanded_conv/depthwise/depthwise + FeatureExtractor/MobilenetV2/expanded_conv/depthwise/Relu6 (CaskConvolution) [TRT] FeatureExtractor/MobilenetV2/expanded_conv/depthwise/depthwise + FeatureExtractor/MobilenetV2/expanded_conv/depthwise/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_medium_nn_v1 [TRT] Tactic: 1062367460111450758 time 14.4434 [TRT] FeatureExtractor/MobilenetV2/expanded_conv/depthwise/depthwise + FeatureExtractor/MobilenetV2/expanded_conv/depthwise/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_large_nn_v1 [TRT] Tactic: 4337000649858996379 time 22.8109 [TRT] FeatureExtractor/MobilenetV2/expanded_conv/depthwise/depthwise + FeatureExtractor/MobilenetV2/expanded_conv/depthwise/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_medium_nn_v1 [TRT] Tactic: 4501471010995462441 time 44.539 [TRT] FeatureExtractor/MobilenetV2/expanded_conv/depthwise/depthwise + FeatureExtractor/MobilenetV2/expanded_conv/depthwise/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_small_nn_v1 [TRT] Tactic: 5137655947464784826 time 21.9238 [TRT] FeatureExtractor/MobilenetV2/expanded_conv/depthwise/depthwise + FeatureExtractor/MobilenetV2/expanded_conv/depthwise/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_medium_nn_v1 [TRT] Tactic: 6645123197870846056 time 22.3931 [TRT] FeatureExtractor/MobilenetV2/expanded_conv/depthwise/depthwise + FeatureExtractor/MobilenetV2/expanded_conv/depthwise/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_large_nn_v1 [TRT] Tactic: -9137461792520977713 time 45.1667 [TRT] FeatureExtractor/MobilenetV2/expanded_conv/depthwise/depthwise + FeatureExtractor/MobilenetV2/expanded_conv/depthwise/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_large_nn_v1 [TRT] Tactic: -6092040395344634144 time 14.796 [TRT] FeatureExtractor/MobilenetV2/expanded_conv/depthwise/depthwise + FeatureExtractor/MobilenetV2/expanded_conv/depthwise/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_small_nn_v1 [TRT] Tactic: -3456450830548107839 time 13.7582 [TRT] FeatureExtractor/MobilenetV2/expanded_conv/depthwise/depthwise + FeatureExtractor/MobilenetV2/expanded_conv/depthwise/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_small_nn_v1 [TRT] Tactic: -410470605513481746 time 43.7929 [TRT] Fastest Tactic: -3456450830548107839 Time: 13.7582 [TRT] --------------- Timing Runner: FeatureExtractor/MobilenetV2/expanded_conv/depthwise/depthwise + FeatureExtractor/MobilenetV2/expanded_conv/depthwise/Relu6 (CudaConvolution) [TRT] Tactic: 0 time 2.0319 [TRT] Tactic: 1 time 12.6501 [TRT] Tactic: 2 time 13.2605 [TRT] Tactic: 6 time 30.5181 [TRT] Fastest Tactic: 0 Time: 2.0319 [TRT] --------------- Timing Runner: FeatureExtractor/MobilenetV2/expanded_conv/depthwise/depthwise + FeatureExtractor/MobilenetV2/expanded_conv/depthwise/Relu6 (CudaDepthwiseConvolution) [TRT] Tactic: -1 time 1.09299 [TRT] Fastest Tactic: -1 Time: 1.09299 [TRT] >>>>>>>>>>>>>>> Chose Runner Type: CudaDepthwiseConvolution Tactic: -1 [TRT]
  [TRT] Autotuning format combination: Half(1,150,22500,720000) -> Half(1,150,22500,720000) [TRT] --------------- Timing Runner: FeatureExtractor/MobilenetV2/expanded_conv/depthwise/depthwise + FeatureExtractor/MobilenetV2/expanded_conv/depthwise/Relu6 (FusedConvActConvolution) [TRT] FusedConvActConvolution has no valid tactics for this config, skipping [TRT] --------------- Timing Runner: FeatureExtractor/MobilenetV2/expanded_conv/depthwise/depthwise + FeatureExtractor/MobilenetV2/expanded_conv/depthwise/Relu6 (CaskConvolution) [TRT] CaskConvolution has no valid tactics for this config, skipping [TRT] --------------- Timing Runner: FeatureExtractor/MobilenetV2/expanded_conv/depthwise/depthwise + FeatureExtractor/MobilenetV2/expanded_conv/depthwise/Relu6 (CudaConvolution) [TRT] Tactic: 0 time 1.78143 [TRT] Tactic: 1 time 0.999843 [TRT] Tactic: 2 time 13.0674 [TRT] Tactic: 6 time 43.3709 [TRT] Fastest Tactic: 1 Time: 0.999843 [TRT] --------------- Timing Runner: FeatureExtractor/MobilenetV2/expanded_conv/depthwise/depthwise + FeatureExtractor/MobilenetV2/expanded_conv/depthwise/Relu6 (CudaDepthwiseConvolution) [TRT] CudaDepthwiseConvolution has no valid tactics for this config, skipping [TRT] >>>>>>>>>>>>>>> Chose Runner Type: CudaConvolution Tactic: 1 [TRT]
  [TRT] --------------- Timing Runner: (Reformat) [TRT] Tactic: 1002 time 0.496016 [TRT] Tactic: 0 time 0.73349 [TRT] Fastest Tactic: 1002 Time: 0.496016 [TRT] --------------- Timing Runner: (Reformat) [TRT] Tactic: 1002 time 1.5131 [TRT] Tactic: 0 time 0.587162 [TRT] Fastest Tactic: 0 Time: 0.587162 [TRT] --------------- Timing Runner: (Reformat) [TRT] Tactic: 1002 time 0.495833 [TRT] Tactic: 0 time 0.622865 [TRT] Fastest Tactic: 1002 Time: 0.495833 [TRT] --------------- Timing Runner: (Reformat) [TRT] Tactic: 1002 time 1.51359 [TRT] Tactic: 0 time 0.57612 [TRT] Fastest Tactic: 0 Time: 0.57612 [TRT] --------------- Timing Runner: (Reformat) [TRT] Tactic: 1002 time 1.96492 [TRT] Tactic: 0 time 0.536041 [TRT] Fastest Tactic: 0 Time: 0.536041 [TRT] --------------- Timing Runner: (Reformat) [TRT] Tactic: 1002 time 1.96664 [TRT] Tactic: 0 time 0.525495 [TRT] Fastest Tactic: 0 Time: 0.525495 [TRT] Autotuning format combination: Float(1,150,22500,720000) -> Float(1,150,22500,360000) [TRT] --------------- Timing Runner: FeatureExtractor/MobilenetV2/expanded_conv/project/Conv2D (LegacySASSConvolution) [TRT] Tactic: 0 time 0.583412 [TRT] Fastest Tactic: 0 Time: 0.583412 [TRT] --------------- Timing Runner: FeatureExtractor/MobilenetV2/expanded_conv/project/Conv2D (FusedConvActConvolution) [TRT] Tactic: 8 time 1.2631 [TRT] Tactic: 9 time 2.4249 [TRT] Tactic: 11 time 1.48753 [TRT] Tactic: 12 time 4.29865 [TRT] Tactic: 17 time 2.18995 [TRT] Tactic: 19 time 2.22443 [TRT] Tactic: 20 time 1.17844 [TRT] Tactic: 21 time 2.375 [TRT] Tactic: 22 time 3.247 [TRT] Tactic: 24 time 1.63497 [TRT] Tactic: 27 time 1.54437 [TRT] Tactic: 28 time 2.77708 [TRT] Tactic: 31 time 2.00862 [TRT] Tactic: 32 time 1.20563 [TRT] Tactic: 34 time 1.13151 [TRT] Tactic: 35 time 1.55143 [TRT] Tactic: 40 time 1.13885 [TRT] Tactic: 46 time 3.04172 [TRT] Tactic: 50 time 3.06333 [TRT] Tactic: 51 time 1.5 [TRT] Tactic: 53 time 1.46833 [TRT] Tactic: 55 time 2.09742 [TRT] Tactic: 59 time 2.24919 [TRT] Tactic: 60 time 1.54326 [TRT] Tactic: 61 time 4.2582 [TRT] Tactic: 63 time 1.46641 [TRT] Tactic: 64 time 1.9362 [TRT] Tactic: 65 time 1.34849 [TRT] Tactic: 68 time 1.15237 [TRT] Tactic: 69 time 1.54841 [TRT] Tactic: 70 time 2.54453 [TRT] Tactic: 74 time 1.56318 [TRT] Tactic: 75 time 1.29628 [TRT] Tactic: 78 time 2.24625 [TRT] Tactic: 79 time 1.63375 [TRT] Tactic: 81 time 2.37049

[dusty-nv]

The SSD models take a few minutes to load the first time (around 5 minutes). I just tried it again here, and it loads and runs as expected on JetPack 4.3.

[Philliec459]

Can I ask what actually is going on the first time you try to run a model vs. every time after that? The first time I tried a few models with my web cam it seemed to go on and on so I killed the program thinking that I had the wrong webcam. Now it works perfect. I would just like to know what the program does the fist time that is do different. Thanks

[WyattAutomation]

Confirmed that this is the case with running the python detectnet-camera.py example on the Nano; both the coco and the SSD models had to run for 8 and 5 minutes respectively on the loop of 'tactic' information that whizzes by on their first load. Subsequent uses start up right away.

Might be worth a quick mention in the documentation for the examples. I did the same thing and fuddled around with swapping/testing cams before I found this thread; wasn't expecting that to be the issue but I'm glad it works and appreciate the help here!