how to support custom layer of PReLU on TensorRT7

./parserHelper.h:99: nvinfer1::DimsCHW parserhelper::getCHW(const nvinfer1::Dims&): Assertion `d.nbDims >= 3’ failed.

Hi,

Based on above error it seems to be dimension issue, dimension of the node should be in NCHW or CHW format.

Also, PRelu is supported in ONNX, please refer below link:
https://github.com/onnx/onnx/blob/master/docs/Operators.md#PRelu

Could you please share more details about the issue or error that you are getting?

Also, can you provide the following information so we can better help?
Provide details on the platforms you are using:
o Linux distro and version
o GPU type
o Nvidia driver version
o CUDA version
o CUDNN version
o Python version [if using python]
o Tensorflow version
o TensorRT version
o If Jetson, OS, hw versions

Also, if possible please share the script and model files.

Thanks

When use trtexec with trt7 to execute https://github.com/CMU-Perceptual-Computing-Lab/openpose/blob/master/models/pose/body_25/pose_deploy.prototxt
and
http://posefs1.perception.cs.cmu.edu/OpenPose/models/pose/body_25/pose_iter_584000.caffemodel
, I can also met the problem

“”./parserHelper.h:99: nvinfer1::DimsCHW parserhelper::getCHW(const nvinfer1::Dims&): Assertion `d.nbDims >= 3’ failed. “”

Hi,

Can you share the trtexec output log in “–verbose” mode?

Thanks

$ /usr/src/tensorrt/bin/trtexec --deploy=pose_deploy.prototxt --model=pose_iter_584000.caffemodel --output=net_output --verbose
&&&& RUNNING TensorRT.trtexec # /usr/src/tensorrt/bin/trtexec --deploy=pose_deploy.prototxt --model=pose_iter_584000.caffemodel --output=net_output --verbose
[02/28/2020-17:01:24] [I] === Model Options ===
[02/28/2020-17:01:24] [I] Format: Caffe
[02/28/2020-17:01:24] [I] Model: pose_iter_584000.caffemodel
[02/28/2020-17:01:24] [I] Prototxt: pose_deploy.prototxt
[02/28/2020-17:01:24] [I] Output: net_output
[02/28/2020-17:01:24] [I] === Build Options ===
[02/28/2020-17:01:24] [I] Max batch: 1
[02/28/2020-17:01:24] [I] Workspace: 16 MB
[02/28/2020-17:01:24] [I] minTiming: 1
[02/28/2020-17:01:24] [I] avgTiming: 8
[02/28/2020-17:01:24] [I] Precision: FP32
[02/28/2020-17:01:24] [I] Calibration:
[02/28/2020-17:01:24] [I] Safe mode: Disabled
[02/28/2020-17:01:24] [I] Save engine:
[02/28/2020-17:01:24] [I] Load engine:
[02/28/2020-17:01:24] [I] Inputs format: fp32:CHW
[02/28/2020-17:01:24] [I] Outputs format: fp32:CHW
[02/28/2020-17:01:24] [I] Input build shapes: model
[02/28/2020-17:01:24] [I] === System Options ===
[02/28/2020-17:01:24] [I] Device: 0
[02/28/2020-17:01:24] [I] DLACore:
[02/28/2020-17:01:24] [I] Plugins:
[02/28/2020-17:01:24] [I] === Inference Options ===
[02/28/2020-17:01:24] [I] Batch: 1
[02/28/2020-17:01:24] [I] Iterations: 10
[02/28/2020-17:01:24] [I] Duration: 3s (+ 200ms warm up)
[02/28/2020-17:01:24] [I] Sleep time: 0ms
[02/28/2020-17:01:24] [I] Streams: 1
[02/28/2020-17:01:24] [I] ExposeDMA: Disabled
[02/28/2020-17:01:24] [I] Spin-wait: Disabled
[02/28/2020-17:01:24] [I] Multithreading: Disabled
[02/28/2020-17:01:24] [I] CUDA Graph: Disabled
[02/28/2020-17:01:24] [I] Skip inference: Disabled
[02/28/2020-17:01:24] [I] Input inference shapes: model
[02/28/2020-17:01:24] [I] Inputs:
[02/28/2020-17:01:24] [I] === Reporting Options ===
[02/28/2020-17:01:24] [I] Verbose: Enabled
[02/28/2020-17:01:24] [I] Averages: 10 inferences
[02/28/2020-17:01:24] [I] Percentile: 99
[02/28/2020-17:01:24] [I] Dump output: Disabled
[02/28/2020-17:01:24] [I] Profile: Disabled
[02/28/2020-17:01:24] [I] Export timing to JSON file:
[02/28/2020-17:01:24] [I] Export output to JSON file:
[02/28/2020-17:01:24] [I] Export profile to JSON file:
[02/28/2020-17:01:24] [I]
[02/28/2020-17:01:24] [V] [TRT] Plugin creator registration succeeded - ::GridAnchor_TRT
[02/28/2020-17:01:24] [V] [TRT] Plugin creator registration succeeded - ::NMS_TRT
[02/28/2020-17:01:24] [V] [TRT] Plugin creator registration succeeded - ::Reorg_TRT
[02/28/2020-17:01:24] [V] [TRT] Plugin creator registration succeeded - ::Region_TRT
[02/28/2020-17:01:24] [V] [TRT] Plugin creator registration succeeded - ::Clip_TRT
[02/28/2020-17:01:24] [V] [TRT] Plugin creator registration succeeded - ::LReLU_TRT
[02/28/2020-17:01:24] [V] [TRT] Plugin creator registration succeeded - ::PriorBox_TRT
[02/28/2020-17:01:24] [V] [TRT] Plugin creator registration succeeded - ::Normalize_TRT
[02/28/2020-17:01:24] [V] [TRT] Plugin creator registration succeeded - ::RPROI_TRT
[02/28/2020-17:01:24] [V] [TRT] Plugin creator registration succeeded - ::BatchedNMS_TRT
[02/28/2020-17:01:24] [V] [TRT] Plugin creator registration succeeded - ::FlattenConcat_TRT
[02/28/2020-17:01:24] [V] [TRT] Plugin creator registration succeeded - ::CropAndResize
[02/28/2020-17:01:24] [V] [TRT] Plugin creator registration succeeded - ::DetectionLayer_TRT
[02/28/2020-17:01:24] [V] [TRT] Plugin creator registration succeeded - ::Proposal
[02/28/2020-17:01:24] [V] [TRT] Plugin creator registration succeeded - ::ProposalLayer_TRT
[02/28/2020-17:01:24] [V] [TRT] Plugin creator registration succeeded - ::PyramidROIAlign_TRT
[02/28/2020-17:01:24] [V] [TRT] Plugin creator registration succeeded - ::ResizeNearest_TRT
[02/28/2020-17:01:24] [V] [TRT] Plugin creator registration succeeded - ::Split
[02/28/2020-17:01:24] [V] [TRT] Plugin creator registration succeeded - ::SpecialSlice_TRT
[02/28/2020-17:01:24] [V] [TRT] Plugin creator registration succeeded - ::InstanceNormalization_TRT
[02/28/2020-17:01:24] [E] [TRT] (Unnamed Layer* 22) [Constant]: constant weights has count 512 but 27 was expected
trtexec: ./parserHelper.h:99: nvinfer1::DimsCHW parserhelper::getCHW(const nvinfer1::Dims&): Assertion `d.nbDims >= 3’ failed.
Aborted (core dumped)

Hi,

Can you try using below command:

trtexec --deploy=pose_deploy.prototxt --output=net_output --verbose

Thanks

&&&& RUNNING TensorRT.trtexec # ./trtexec --deploy=models/resnet101.prototxt --output=fc1 --verbose
[03/25/2020-17:13:32] [I] === Model Options ===
[03/25/2020-17:13:32] [I] Format: Caffe
[03/25/2020-17:13:32] [I] Model:
[03/25/2020-17:13:32] [I] Prototxt: models/resnet101.prototxt
[03/25/2020-17:13:32] [I] Output: fc1
[03/25/2020-17:13:32] [I] === Build Options ===
[03/25/2020-17:13:32] [I] Max batch: 1
[03/25/2020-17:13:32] [I] Workspace: 16 MB
[03/25/2020-17:13:32] [I] minTiming: 1
[03/25/2020-17:13:32] [I] avgTiming: 8
[03/25/2020-17:13:32] [I] Precision: FP32
[03/25/2020-17:13:32] [I] Calibration:
[03/25/2020-17:13:32] [I] Safe mode: Disabled
[03/25/2020-17:13:32] [I] Save engine:
[03/25/2020-17:13:32] [I] Load engine:
[03/25/2020-17:13:32] [I] Inputs format: fp32:CHW
[03/25/2020-17:13:32] [I] Outputs format: fp32:CHW
[03/25/2020-17:13:32] [I] Input build shapes: model
[03/25/2020-17:13:32] [I] === System Options ===
[03/25/2020-17:13:32] [I] Device: 0
[03/25/2020-17:13:32] [I] DLACore:
[03/25/2020-17:13:32] [I] Plugins:
[03/25/2020-17:13:32] [I] === Inference Options ===
[03/25/2020-17:13:32] [I] Batch: 1
[03/25/2020-17:13:32] [I] Iterations: 10
[03/25/2020-17:13:32] [I] Duration: 3s (+ 200ms warm up)
[03/25/2020-17:13:32] [I] Sleep time: 0ms
[03/25/2020-17:13:32] [I] Streams: 1
[03/25/2020-17:13:32] [I] ExposeDMA: Disabled
[03/25/2020-17:13:32] [I] Spin-wait: Disabled
[03/25/2020-17:13:32] [I] Multithreading: Disabled
[03/25/2020-17:13:32] [I] CUDA Graph: Disabled
[03/25/2020-17:13:32] [I] Skip inference: Disabled
[03/25/2020-17:13:32] [I] Input inference shapes: model
[03/25/2020-17:13:32] [I] Inputs:
[03/25/2020-17:13:32] [I] === Reporting Options ===
[03/25/2020-17:13:32] [I] Verbose: Enabled
[03/25/2020-17:13:32] [I] Averages: 10 inferences
[03/25/2020-17:13:32] [I] Percentile: 99
[03/25/2020-17:13:32] [I] Dump output: Disabled
[03/25/2020-17:13:32] [I] Profile: Disabled
[03/25/2020-17:13:32] [I] Export timing to JSON file:
[03/25/2020-17:13:32] [I] Export output to JSON file:
[03/25/2020-17:13:32] [I] Export profile to JSON file:
[03/25/2020-17:13:32] [I]
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::GridAnchor_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::NMS_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::Reorg_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::Region_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::Clip_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::LReLU_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::NMS_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::Reorg_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::Region_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::Clip_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::LReLU_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::PriorBox_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::Normalize_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::RPROI_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::BatchedNMS_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::FlattenConcat_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::CropAndResize
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::DetectionLayer_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::Proposal
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::ProposalLayer_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::PyramidROIAlign_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::ResizeNearest_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::Split
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::SpecialSlice_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::InstanceNormalization_TRT
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:34] [V] [TRT] Applying generic optimizations to the graph for inference.
[03/25/2020-17:13:34] [V] [TRT] Original: 561 layers
[03/25/2020-17:13:34] [V] [TRT] After dead-layer removal: 561 layers
[03/25/2020-17:13:34] [V] [TRT] After Myelin optimization: 561 layers
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from conv0 with scale bn0
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from conv0 with scale bn0_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit1_conv1 with scale stage1_unit1_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit1_conv1 with scale stage1_unit1_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit1_conv2 with scale stage1_unit1_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit1_conv2 with scale stage1_unit1_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit1_conv1sc with scale stage1_unit1_sc
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit1_conv1sc with scale stage1_unit1_sc_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit2_conv1 with scale stage1_unit2_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit2_conv1 with scale stage1_unit2_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit2_conv2 with scale stage1_unit2_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit2_conv2 with scale stage1_unit2_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit3_conv1 with scale stage1_unit3_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit3_conv1 with scale stage1_unit3_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit3_conv2 with scale stage1_unit3_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit3_conv2 with scale stage1_unit3_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit1_conv1 with scale stage2_unit1_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit1_conv1 with scale stage2_unit1_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit1_conv2 with scale stage2_unit1_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit1_conv2 with scale stage2_unit1_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit1_conv1sc with scale stage2_unit1_sc
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit1_conv1sc with scale stage2_unit1_sc_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit2_conv1 with scale stage2_unit2_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit2_conv1 with scale stage2_unit2_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit2_conv2 with scale stage2_unit2_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit2_conv2 with scale stage2_unit2_bn3_scale
.
.
.
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04823 ms - Host latency: 5.07889 ms (end to end 9.9121 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04474 ms - Host latency: 5.07551 ms (end to end 9.9084 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04586 ms - Host latency: 5.07714 ms (end to end 9.9125 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04904 ms - Host latency: 5.07987 ms (end to end 9.9126 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04705 ms - Host latency: 5.07745 ms (end to end 9.90737 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04814 ms - Host latency: 5.07946 ms (end to end 9.91608 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04509 ms - Host latency: 5.07633 ms (end to end 9.91143 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.0468 ms - Host latency: 5.07793 ms (end to end 9.91688 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04841 ms - Host latency: 5.07917 ms (end to end 9.91551 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04529 ms - Host latency: 5.07672 ms (end to end 9.90726 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.0467 ms - Host latency: 5.07747 ms (end to end 9.91728 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04617 ms - Host latency: 5.07726 ms (end to end 9.90934 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.05204 ms - Host latency: 5.08281 ms (end to end 9.91638 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04856 ms - Host latency: 5.07975 ms (end to end 9.91348 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04738 ms - Host latency: 5.07809 ms (end to end 9.91643 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04827 ms - Host latency: 5.07969 ms (end to end 9.91326 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04692 ms - Host latency: 5.07737 ms (end to end 9.90974 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04641 ms - Host latency: 5.07761 ms (end to end 9.91387 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.0491 ms - Host latency: 5.07952 ms (end to end 9.91646 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04622 ms - Host latency: 5.08196 ms (end to end 9.9386 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.05193 ms - Host latency: 5.08367 ms (end to end 9.94683 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04937 ms - Host latency: 5.08359 ms (end to end 9.95549 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.05305 ms - Host latency: 5.0855 ms (end to end 9.95464 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.051 ms - Host latency: 5.08147 ms (end to end 9.94031 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04595 ms - Host latency: 5.07942 ms (end to end 9.9156 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04983 ms - Host latency: 5.08032 ms (end to end 9.91716 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04668 ms - Host latency: 5.07778 ms (end to end 9.90874 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04858 ms - Host latency: 5.08 ms (end to end 9.91248 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04958 ms - Host latency: 5.07969 ms (end to end 9.91143 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04502 ms - Host latency: 5.07627 ms (end to end 9.90671 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04485 ms - Host latency: 5.07593 ms (end to end 9.90537 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04771 ms - Host latency: 5.07876 ms (end to end 9.91702 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04717 ms - Host latency: 5.07881 ms (end to end 9.91228 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04299 ms - Host latency: 5.07373 ms (end to end 9.90508 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04873 ms - Host latency: 5.08101 ms (end to end 9.90645 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.05 ms - Host latency: 5.08298 ms (end to end 9.8959 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04712 ms - Host latency: 5.07808 ms (end to end 9.89204 ms)
[03/25/2020-17:14:08] [I] Host latency
[03/25/2020-17:14:08] [I] min: 5.04639 ms (end to end 5.33325 ms)
[03/25/2020-17:14:08] [I] max: 6.17398 ms (end to end 12.0827 ms)
[03/25/2020-17:14:08] [I] mean: 5.14821 ms (end to end 10.0103 ms)
[03/25/2020-17:14:08] [I] median: 5.08014 ms (end to end 9.91394 ms)
[03/25/2020-17:14:08] [I] percentile: 6.16275 ms at 99% (end to end 12.0636 ms at 99%)
[03/25/2020-17:14:08] [I] throughput: 195.019 qps
[03/25/2020-17:14:08] [I] walltime: 3.00996 s
[03/25/2020-17:14:08] [I] GPU Compute
[03/25/2020-17:14:08] [I] min: 5.01349 ms
[03/25/2020-17:14:08] [I] max: 6.14093 ms
[03/25/2020-17:14:08] [I] mean: 5.11657 ms
[03/25/2020-17:14:08] [I] median: 5.04932 ms
[03/25/2020-17:14:08] [I] percentile: 6.1317 ms at 99%
[03/25/2020-17:14:08] [I] total compute time: 3.00343 s

I also occured this problem when parse resnet_101 network. ./parserHelper.h:99: nvinfer1::DimsCHW parserhelper::getCHW(const nvinfer1::Dims&): Assertion `d.nbDims >= 3’ failed. “”
Here is the log.

&&&& RUNNING TensorRT.trtexec # ./trtexec --deploy=models/resnet101.prototxt --output=fc1 --verbose
[03/25/2020-17:13:32] [I] === Model Options ===
[03/25/2020-17:13:32] [I] Format: Caffe
[03/25/2020-17:13:32] [I] Model:
[03/25/2020-17:13:32] [I] Prototxt: models/resnet101.prototxt
[03/25/2020-17:13:32] [I] Output: fc1
[03/25/2020-17:13:32] [I] === Build Options ===
[03/25/2020-17:13:32] [I] Max batch: 1
[03/25/2020-17:13:32] [I] Workspace: 16 MB
[03/25/2020-17:13:32] [I] minTiming: 1
[03/25/2020-17:13:32] [I] avgTiming: 8
[03/25/2020-17:13:32] [I] Precision: FP32
[03/25/2020-17:13:32] [I] Calibration:
[03/25/2020-17:13:32] [I] Safe mode: Disabled
[03/25/2020-17:13:32] [I] Save engine:
[03/25/2020-17:13:32] [I] Load engine:
[03/25/2020-17:13:32] [I] Inputs format: fp32:CHW
[03/25/2020-17:13:32] [I] Outputs format: fp32:CHW
[03/25/2020-17:13:32] [I] Input build shapes: model
[03/25/2020-17:13:32] [I] === System Options ===
[03/25/2020-17:13:32] [I] Device: 0
[03/25/2020-17:13:32] [I] DLACore:
[03/25/2020-17:13:32] [I] Plugins:
[03/25/2020-17:13:32] [I] === Inference Options ===
[03/25/2020-17:13:32] [I] Batch: 1
[03/25/2020-17:13:32] [I] Iterations: 10
[03/25/2020-17:13:32] [I] Duration: 3s (+ 200ms warm up)
[03/25/2020-17:13:32] [I] Sleep time: 0ms
[03/25/2020-17:13:32] [I] Streams: 1
[03/25/2020-17:13:32] [I] ExposeDMA: Disabled
[03/25/2020-17:13:32] [I] Spin-wait: Disabled
[03/25/2020-17:13:32] [I] Multithreading: Disabled
[03/25/2020-17:13:32] [I] CUDA Graph: Disabled
[03/25/2020-17:13:32] [I] Skip inference: Disabled
[03/25/2020-17:13:32] [I] Input inference shapes: model
[03/25/2020-17:13:32] [I] Inputs:
[03/25/2020-17:13:32] [I] === Reporting Options ===
[03/25/2020-17:13:32] [I] Verbose: Enabled
[03/25/2020-17:13:32] [I] Averages: 10 inferences
[03/25/2020-17:13:32] [I] Percentile: 99
[03/25/2020-17:13:32] [I] Dump output: Disabled
[03/25/2020-17:13:32] [I] Profile: Disabled
[03/25/2020-17:13:32] [I] Export timing to JSON file:
[03/25/2020-17:13:32] [I] Export output to JSON file:
[03/25/2020-17:13:32] [I] Export profile to JSON file:
[03/25/2020-17:13:32] [I]
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::GridAnchor_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::NMS_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::Reorg_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::Region_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::Clip_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::LReLU_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::Clip_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::LReLU_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::PriorBox_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::Normalize_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::RPROI_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::BatchedNMS_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::FlattenConcat_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::CropAndResize
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::DetectionLayer_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::Proposal
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::ProposalLayer_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::PyramidROIAlign_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::ResizeNearest_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::Split
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::SpecialSlice_TRT
[03/25/2020-17:13:32] [V] [TRT] Plugin creator registration succeeded - ::InstanceNormalization_TRT
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:32] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:33] [V] [TRT] Bias weights are not set yet. Bias weights can be set using setInput(2, bias_tensor) API call.
[03/25/2020-17:13:34] [V] [TRT] Applying generic optimizations to the graph for inference.
[03/25/2020-17:13:34] [V] [TRT] Original: 561 layers
[03/25/2020-17:13:34] [V] [TRT] After dead-layer removal: 561 layers
[03/25/2020-17:13:34] [V] [TRT] After Myelin optimization: 561 layers
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from conv0 with scale bn0
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from conv0 with scale bn0_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit1_conv1 with scale stage1_unit1_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit1_conv1 with scale stage1_unit1_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit1_conv2 with scale stage1_unit1_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit1_conv2 with scale stage1_unit1_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit1_conv1sc with scale stage1_unit1_sc
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit1_conv1sc with scale stage1_unit1_sc_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit2_conv1 with scale stage1_unit2_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit2_conv1 with scale stage1_unit2_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit2_conv2 with scale stage1_unit2_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit2_conv2 with scale stage1_unit2_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit3_conv1 with scale stage1_unit3_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit3_conv1 with scale stage1_unit3_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit3_conv2 with scale stage1_unit3_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage1_unit3_conv2 with scale stage1_unit3_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit1_conv1 with scale stage2_unit1_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit1_conv1 with scale stage2_unit1_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit1_conv2 with scale stage2_unit1_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit1_conv2 with scale stage2_unit1_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit1_conv1sc with scale stage2_unit1_sc
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit1_conv1sc with scale stage2_unit1_sc_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit2_conv1 with scale stage2_unit2_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit2_conv1 with scale stage2_unit2_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit2_conv2 with scale stage2_unit2_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit2_conv2 with scale stage2_unit2_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit3_conv1 with scale stage2_unit3_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit3_conv1 with scale stage2_unit3_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit3_conv2 with scale stage2_unit3_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit3_conv1 with scale stage2_unit3_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit3_conv2 with scale stage2_unit3_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit3_conv2 with scale stage2_unit3_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit4_conv1 with scale stage2_unit4_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit4_conv1 with scale stage2_unit4_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit4_conv2 with scale stage2_unit4_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit4_conv2 with scale stage2_unit4_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit5_conv1 with scale stage2_unit5_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit5_conv1 with scale stage2_unit5_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit5_conv2 with scale stage2_unit5_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit5_conv2 with scale stage2_unit5_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit6_conv1 with scale stage2_unit6_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit6_conv1 with scale stage2_unit6_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit6_conv2 with scale stage2_unit6_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit6_conv2 with scale stage2_unit6_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit7_conv1 with scale stage2_unit7_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit7_conv1 with scale stage2_unit7_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit7_conv2 with scale stage2_unit7_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit7_conv2 with scale stage2_unit7_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit8_conv1 with scale stage2_unit8_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit8_conv1 with scale stage2_unit8_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit8_conv2 with scale stage2_unit8_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit8_conv2 with scale stage2_unit8_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit9_conv1 with scale stage2_unit9_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit9_conv1 with scale stage2_unit9_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit9_conv2 with scale stage2_unit9_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit9_conv2 with scale stage2_unit9_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit10_conv1 with scale stage2_unit10_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit10_conv1 with scale stage2_unit10_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit10_conv2 with scale stage2_unit10_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit10_conv2 with scale stage2_unit10_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit11_conv1 with scale stage2_unit11_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit11_conv1 with scale stage2_unit11_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit11_conv2 with scale stage2_unit11_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit11_conv2 with scale stage2_unit11_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit12_conv1 with scale stage2_unit12_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit12_conv1 with scale stage2_unit12_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit12_conv2 with scale stage2_unit12_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit12_conv2 with scale stage2_unit12_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit13_conv1 with scale stage2_unit13_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit13_conv1 with scale stage2_unit13_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit13_conv2 with scale stage2_unit13_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage2_unit13_conv2 with scale stage2_unit13_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit1_conv1 with scale stage3_unit1_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit1_conv1 with scale stage3_unit1_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit1_conv2 with scale stage3_unit1_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit1_conv2 with scale stage3_unit1_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit1_conv1sc with scale stage3_unit1_sc
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit1_conv1sc with scale stage3_unit1_sc_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit2_conv1 with scale stage3_unit2_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit2_conv1 with scale stage3_unit2_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit2_conv2 with scale stage3_unit2_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit2_conv1 with scale stage3_unit2_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit2_conv2 with scale stage3_unit2_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit2_conv2 with scale stage3_unit2_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit3_conv1 with scale stage3_unit3_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit3_conv1 with scale stage3_unit3_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit3_conv2 with scale stage3_unit3_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit3_conv2 with scale stage3_unit3_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit4_conv1 with scale stage3_unit4_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit4_conv1 with scale stage3_unit4_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit4_conv2 with scale stage3_unit4_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit4_conv2 with scale stage3_unit4_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit5_conv1 with scale stage3_unit5_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit5_conv1 with scale stage3_unit5_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit5_conv2 with scale stage3_unit5_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit5_conv2 with scale stage3_unit5_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit6_conv1 with scale stage3_unit6_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit6_conv1 with scale stage3_unit6_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit6_conv2 with scale stage3_unit6_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit6_conv2 with scale stage3_unit6_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit7_conv1 with scale stage3_unit7_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit7_conv1 with scale stage3_unit7_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit7_conv2 with scale stage3_unit7_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit7_conv2 with scale stage3_unit7_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit8_conv1 with scale stage3_unit8_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit8_conv1 with scale stage3_unit8_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit8_conv2 with scale stage3_unit8_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit8_conv2 with scale stage3_unit8_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit9_conv1 with scale stage3_unit9_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit9_conv1 with scale stage3_unit9_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit9_conv2 with scale stage3_unit9_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit9_conv2 with scale stage3_unit9_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit10_conv1 with scale stage3_unit10_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit10_conv1 with scale stage3_unit10_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit10_conv2 with scale stage3_unit10_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit10_conv2 with scale stage3_unit10_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit11_conv1 with scale stage3_unit11_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit11_conv1 with scale stage3_unit11_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit11_conv2 with scale stage3_unit11_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit11_conv2 with scale stage3_unit11_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit12_conv1 with scale stage3_unit12_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit12_conv1 with scale stage3_unit12_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit12_conv2 with scale stage3_unit12_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit12_conv2 with scale stage3_unit12_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit13_conv1 with scale stage3_unit13_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit13_conv1 with scale stage3_unit13_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit13_conv2 with scale stage3_unit13_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit13_conv2 with scale stage3_unit13_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit14_conv1 with scale stage3_unit14_bn2
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit14_conv1 with scale stage3_unit14_bn2_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit14_conv2 with scale stage3_unit14_bn3
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit14_conv2 with scale stage3_unit14_bn3_scale
[03/25/2020-17:13:34] [V] [TRT] Fusing convolution weights from stage3_unit15_conv1 with scale stage3_unit15_bn2
.
.
.
[03/25/2020-17:14:04] [V] [TRT] stage2_unit4_conv2 + _plus6 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage2_unit5_conv1 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage2_unit5_conv2 + _plus7 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage2_unit6_conv1 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage2_unit6_conv2 + _plus8 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage2_unit7_conv1 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage2_unit7_conv2 + _plus9 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage2_unit8_conv1 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage2_unit8_conv2 + _plus10 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage2_unit9_conv1 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage2_unit9_conv2 + _plus11 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage2_unit10_conv1 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage2_unit10_conv2 + _plus12 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage2_unit11_conv1 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage2_unit11_conv2 + _plus13 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage2_unit12_conv1 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage2_unit12_conv2 + _plus14 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage2_unit13_conv1 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage2_unit13_conv2 + _plus15 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit1_conv1 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit1_conv1sc (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_small_nn_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit1_conv2 + _plus16 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_small_nn_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit3_conv2 + _plus18 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit5_conv1 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit6_conv1 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit8_conv1 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit9_conv1 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit10_conv2 + _plus25 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit11_conv2 + _plus26 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit14_conv2 + _plus29 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit15_conv2 + _plus30 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit16_conv2 + _plus31 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit15_conv2 + _plus30 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit16_conv2 + _plus31 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit17_conv1 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit18_conv1 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit18_conv2 + _plus33 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit19_conv1 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit19_conv2 + _plus34 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit20_conv2 + _plus35 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit24_conv2 + _plus39 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit25_conv2 + _plus40 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit26_conv2 + _plus41 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit27_conv1 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit27_conv2 + _plus42 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit28_conv2 + _plus43 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit29_conv1 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit29_conv2 + _plus44 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage3_unit30_conv2 + _plus45 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage4_unit1_conv1 (scudnn_winograd) Set Tactic Name: maxwell_scudnn_winograd_128x128_ldg1_ldg4_relu_tile148n_nt_v1
[03/25/2020-17:14:04] [V] [TRT] stage4_unit1_conv1sc (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_small_nn_v1
[03/25/2020-17:14:04] [V] [TRT] Engine generation completed in 30.3943 seconds.
[03/25/2020-17:14:04] [V] [TRT] Engine Layer Information:
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomImplicitGemm): conv0, Tactic: 0, data[Float(3,112,112)] -> bn0[Float(64,112,112)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 3) [Constant], Tactic: 0, -> (Unnamed Layer* 3) [Constant]_output[Float(1,112,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): relu0, Tactic: 256, bn0[Float(64,112,112)], (Unnamed Layer* 3) [Constant]_output[Float(1,112,1)] -> relu0[Float(64,112,112)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage1_unit1_bn1 + stage1_unit1_bn1_scale, Tactic: 0, relu0[Float(64,112,112)] -> stage1_unit1_bn1[Float(64,112,112)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage1_unit1_conv1, Tactic: 3827454225649558724, stage1_unit1_bn1[Float(64,112,112)] -> stage1_unit1_bn2[Float(64,112,112)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 10) [Constant], Tactic: 0, -> (Unnamed Layer* 10) [Constant]_output[Float(1,112,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage1_unit1_relu1, Tactic: 256, stage1_unit1_bn2[Float(64,112,112)], (Unnamed Layer* 10) [Constant]_output[Float(1,112,1)] -> stage1_unit1_relu1[Float(64,112,112)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn): stage1_unit1_conv1sc, Tactic: -37215280111360163, relu0[Float(64,112,112)] -> stage1_unit1_sc[Float(64,56,56)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn): stage1_unit1_conv2 + _plus0, Tactic: 5137655947464784826, stage1_unit1_relu1[Float(64,112,112)], stage1_unit1_sc[Float(64,56,56)] -> _plus0[Float(64,56,56)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage1_unit2_bn1 + stage1_unit2_bn1_scale, Tactic: 0, _plus0[Float(64,56,56)] -> stage1_unit2_bn1[Float(64,56,56)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage1_unit2_conv1, Tactic: 3827454225649558724, stage1_unit2_bn1[Float(64,56,56)] -> stage1_unit2_bn2[Float(64,56,56)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 24) [Constant], Tactic: 0, -> (Unnamed Layer* 24) [Constant]_output[Float(1,56,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage1_unit2_relu1, Tactic: 512, stage1_unit2_bn2[Float(64,56,56)], (Unnamed Layer* 24) [Constant]_output[Float(1,56,1)] -> stage1_unit2_relu1[Float(64,56,56)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage1_unit2_conv2 + _plus1, Tactic: 3827454225649558724, stage1_unit2_relu1[Float(64,56,56)], _plus0[Float(64,56,56)] -> _plus1[Float(64,56,56)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage1_unit3_bn1 + stage1_unit3_bn1_scale, Tactic: 0, _plus1[Float(64,56,56)] -> stage1_unit3_bn1[Float(64,56,56)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage1_unit3_conv1, Tactic: 3827454225649558724, stage1_unit3_bn1[Float(64,56,56)] -> stage1_unit3_bn2[Float(64,56,56)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 35) [Constant], Tactic: 0, -> (Unnamed Layer* 35) [Constant]_output[Float(1,56,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage1_unit3_relu1, Tactic: 256, stage1_unit3_bn2[Float(64,56,56)], (Unnamed Layer* 35) [Constant]_output[Float(1,56,1)] -> stage1_unit3_relu1[Float(64,56,56)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage1_unit3_conv2 + _plus2, Tactic: 3827454225649558724, stage1_unit3_relu1[Float(64,56,56)], _plus1[Float(64,56,56)] -> _plus2[Float(64,56,56)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage2_unit1_bn1 + stage2_unit1_bn1_scale, Tactic: 0, _plus2[Float(64,56,56)] -> stage2_unit1_bn1[Float(64,56,56)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage2_unit1_conv1, Tactic: 3827454225649558724, stage2_unit1_bn1[Float(64,56,56)] -> stage2_unit1_bn2[Float(128,56,56)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 46) [Constant], Tactic: 0, -> (Unnamed Layer* 46) [Constant]_output[Float(1,56,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage2_unit1_relu1, Tactic: 512, stage2_unit1_bn2[Float(128,56,56)], (Unnamed Layer* 46) [Constant]_output[Float(1,56,1)] -> stage2_unit1_relu1[Float(128,56,56)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn): stage2_unit1_conv1sc, Tactic: 6645123197870846056, _plus2[Float(64,56,56)] -> stage2_unit1_sc[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn): stage2_unit1_conv2 + _plus3, Tactic: 5137655947464784826, stage2_unit1_relu1[Float(128,56,56)], stage2_unit1_sc[Float(128,28,28)] -> _plus3[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage2_unit2_bn1 + stage2_unit2_bn1_scale, Tactic: 0, _plus3[Float(128,28,28)] -> stage2_unit2_bn1[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage2_unit2_conv1, Tactic: 3827454225649558724, stage2_unit2_bn1[Float(128,28,28)] -> stage2_unit2_bn2[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 60) [Constant], Tactic: 0, -> (Unnamed Layer* 60) [Constant]_output[Float(1,28,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage2_unit2_relu1, Tactic: 128, stage2_unit2_bn2[Float(128,28,28)], (Unnamed Layer* 60) [Constant]_output[Float(1,28,1)] -> stage2_unit2_relu1[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage2_unit2_conv2 + _plus4, Tactic: 3827454225649558724, stage2_unit2_relu1[Float(128,28,28)], _plus3[Float(128,28,28)] -> _plus4[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage2_unit3_bn1 + stage2_unit3_bn1_scale, Tactic: 0, _plus4[Float(128,28,28)] -> stage2_unit3_bn1[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage2_unit2_conv2 + _plus4, Tactic: 3827454225649558724, stage2_unit2_relu1[Float(128,28,28)], _plus3[Float(128,28,28)] -> _plus4[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage2_unit3_bn1 + stage2_unit3_bn1_scale, Tactic: 0, _plus4[Float(128,28,28)] -> stage2_unit3_bn1[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage2_unit3_conv1, Tactic: 3827454225649558724, stage2_unit3_bn1[Float(128,28,28)] -> stage2_unit3_bn2[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 71) [Constant], Tactic: 0, -> (Unnamed Layer* 71) [Constant]_output[Float(1,28,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage2_unit3_relu1, Tactic: 128, stage2_unit3_bn2[Float(128,28,28)], (Unnamed Layer* 71) [Constant]_output[Float(1,28,1)] -> stage2_unit3_relu1[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage2_unit3_conv2 + _plus5, Tactic: 3827454225649558724, stage2_unit3_relu1[Float(128,28,28)], _plus4[Float(128,28,28)] -> _plus5[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage2_unit4_bn1 + stage2_unit4_bn1_scale, Tactic: 0, _plus5[Float(128,28,28)] -> stage2_unit4_bn1[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage2_unit4_conv1, Tactic: 3827454225649558724, stage2_unit4_bn1[Float(128,28,28)] -> stage2_unit4_bn2[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 82) [Constant], Tactic: 0, -> (Unnamed Layer* 82) [Constant]_output[Float(1,28,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage2_unit4_relu1, Tactic: 256, stage2_unit4_bn2[Float(128,28,28)], (Unnamed Layer* 82) [Constant]_output[Float(1,28,1)] -> stage2_unit4_relu1[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage2_unit4_conv2 + _plus6, Tactic: 3827454225649558724, stage2_unit4_relu1[Float(128,28,28)], _plus5[Float(128,28,28)] -> _plus6[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage2_unit5_bn1 + stage2_unit5_bn1_scale, Tactic: 0, _plus6[Float(128,28,28)] -> stage2_unit5_bn1[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage2_unit5_conv1, Tactic: 3827454225649558724, stage2_unit5_bn1[Float(128,28,28)] -> stage2_unit5_bn2[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 93) [Constant], Tactic: 0, -> (Unnamed Layer* 93) [Constant]_output[Float(1,28,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage2_unit5_relu1, Tactic: 512, stage2_unit5_bn2[Float(128,28,28)], (Unnamed Layer* 93) [Constant]_output[Float(1,28,1)] -> stage2_unit5_relu1[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage2_unit5_conv2 + _plus7, Tactic: 3827454225649558724, stage2_unit5_relu1[Float(128,28,28)], _plus6[Float(128,28,28)] -> _plus7[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage2_unit6_bn1 + stage2_unit6_bn1_scale, Tactic: 0, _plus7[Float(128,28,28)] -> stage2_unit6_bn1[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage2_unit6_conv1, Tactic: 3827454225649558724, stage2_unit6_bn1[Float(128,28,28)] -> stage2_unit6_bn2[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 104) [Constant], Tactic: 0, -> (Unnamed Layer* 104) [Constant]_output[Float(1,28,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage2_unit6_relu1, Tactic: 512, stage2_unit6_bn2[Float(128,28,28)], (Unnamed Layer* 104) [Constant]_output[Float(1,28,1)] -> stage2_unit6_relu1[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage2_unit6_conv2 + _plus8, Tactic: 3827454225649558724, stage2_unit6_relu1[Float(128,28,28)], _plus7[Float(128,28,28)] -> _plus8[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage2_unit7_bn1 + stage2_unit7_bn1_scale, Tactic: 0, _plus8[Float(128,28,28)] -> stage2_unit7_bn1[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage2_unit7_conv1, Tactic: 3827454225649558724, stage2_unit7_bn1[Float(128,28,28)] -> stage2_unit7_bn2[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 115) [Constant], Tactic: 0, -> (Unnamed Layer* 115) [Constant]_output[Float(1,28,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage2_unit7_relu1, Tactic: 256, stage2_unit7_bn2[Float(128,28,28)], (Unnamed Layer* 115) [Constant]_output[Float(1,28,1)] -> stage2_unit7_relu1[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage2_unit7_conv2 + _plus9, Tactic: 3827454225649558724, stage2_unit7_relu1[Float(128,28,28)], _plus8[Float(128,28,28)] -> _plus9[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage2_unit8_bn1 + stage2_unit8_bn1_scale, Tactic: 0, _plus9[Float(128,28,28)] -> stage2_unit8_bn1[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage2_unit8_conv1, Tactic: 3827454225649558724, stage2_unit8_bn1[Float(128,28,28)] -> stage2_unit8_bn2[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 126) [Constant], Tactic: 0, -> (Unnamed Layer* 126) [Constant]_output[Float(1,28,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage2_unit8_relu1, Tactic: 128, stage2_unit8_bn2[Float(128,28,28)], (Unnamed Layer* 126) [Constant]_output[Float(1,28,1)] -> stage2_unit8_relu1[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage2_unit8_conv2 + _plus10, Tactic: 3827454225649558724, stage2_unit8_relu1[Float(128,28,28)], _plus9[Float(128,28,28)] -> _plus10[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage2_unit9_bn1 + stage2_unit9_bn1_scale, Tactic: 0, _plus10[Float(128,28,28)] -> stage2_unit9_bn1[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage2_unit9_conv1, Tactic: 3827454225649558724, stage2_unit9_bn1[Float(128,28,28)] -> stage2_unit9_bn2[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 137) [Constant], Tactic: 0, -> (Unnamed Layer* 137) [Constant]_output[Float(1,28,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage2_unit9_relu1, Tactic: 128, stage2_unit9_bn2[Float(128,28,28)], (Unnamed Layer* 137) [Constant]_output[Float(1,28,1)] -> stage2_unit9_relu1[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage2_unit9_conv2 + _plus11, Tactic: 3827454225649558724, stage2_unit9_relu1[Float(128,28,28)], _plus10[Float(128,28,28)] -> _plus11[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage2_unit10_bn1 + stage2_unit10_bn1_scale, Tactic: 0, _plus11[Float(128,28,28)] -> stage2_unit10_bn1[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage2_unit10_conv1, Tactic: 3827454225649558724, stage2_unit10_bn1[Float(128,28,28)] -> stage2_unit10_bn2[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 148) [Constant], Tactic: 0, -> (Unnamed Layer* 148) [Constant]_output[Float(1,28,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage2_unit10_relu1, Tactic: 128, stage2_unit10_bn2[Float(128,28,28)], (Unnamed Layer* 148) [Constant]_output[Float(1,28,1)] -> stage2_unit10_relu1[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage2_unit10_conv2 + _plus12, Tactic: 3827454225649558724, stage2_unit10_relu1[Float(128,28,28)], _plus11[Float(128,28,28)] -> _plus12[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage2_unit11_bn1 + stage2_unit11_bn1_scale, Tactic: 0, _plus12[Float(128,28,28)] -> stage2_unit11_bn1[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage2_unit11_conv1, Tactic: 3827454225649558724, stage2_unit11_bn1[Float(128,28,28)] -> stage2_unit11_bn2[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 159) [Constant], Tactic: 0, -> (Unnamed Layer* 159) [Constant]_output[Float(1,28,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage2_unit11_relu1, Tactic: 128, stage2_unit11_bn2[Float(128,28,28)], (Unnamed Layer* 159) [Constant]_output[Float(1,28,1)] -> stage2_unit11_relu1[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage2_unit11_conv2 + _plus13, Tactic: 3827454225649558724, stage2_unit11_relu1[Float(128,28,28)], _plus12[Float(128,28,28)] -> _plus13[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage2_unit12_bn1 + stage2_unit12_bn1_scale, Tactic: 0, _plus13[Float(128,28,28)] -> stage2_unit12_bn1[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage2_unit12_conv1, Tactic: 3827454225649558724, stage2_unit12_bn1[Float(128,28,28)] -> stage2_unit12_bn2[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 170) [Constant], Tactic: 0, -> (Unnamed Layer* 170) [Constant]_output[Float(1,28,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage2_unit12_relu1, Tactic: 128, stage2_unit12_bn2[Float(128,28,28)], (Unnamed Layer* 170) [Constant]_output[Float(1,28,1)] -> stage2_unit12_relu1[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage2_unit12_conv2 + _plus14, Tactic: 3827454225649558724, stage2_unit12_relu1[Float(128,28,28)], _plus13[Float(128,28,28)] -> _plus14[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage2_unit13_bn1 + stage2_unit13_bn1_scale, Tactic: 0, _plus14[Float(128,28,28)] -> stage2_unit13_bn1[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage2_unit12_conv2 + _plus14, Tactic: 3827454225649558724, stage2_unit12_relu1[Float(128,28,28)], _plus13[Float(128,28,28)] -> _plus14[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage2_unit13_bn1 + stage2_unit13_bn1_scale, Tactic: 0, _plus14[Float(128,28,28)] -> stage2_unit13_bn1[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage2_unit13_conv1, Tactic: 3827454225649558724, stage2_unit13_bn1[Float(128,28,28)] -> stage2_unit13_bn2[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 181) [Constant], Tactic: 0, -> (Unnamed Layer* 181) [Constant]_output[Float(1,28,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage2_unit13_relu1, Tactic: 128, stage2_unit13_bn2[Float(128,28,28)], (Unnamed Layer* 181) [Constant]_output[Float(1,28,1)] -> stage2_unit13_relu1[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage2_unit13_conv2 + _plus15, Tactic: 3827454225649558724, stage2_unit13_relu1[Float(128,28,28)], _plus14[Float(128,28,28)] -> _plus15[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit1_bn1 + stage3_unit1_bn1_scale, Tactic: 0, _plus15[Float(128,28,28)] -> stage3_unit1_bn1[Float(128,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit1_conv1, Tactic: 3827454225649558724, stage3_unit1_bn1[Float(128,28,28)] -> stage3_unit1_bn2[Float(256,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 192) [Constant], Tactic: 0, -> (Unnamed Layer* 192) [Constant]_output[Float(1,28,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit1_relu1, Tactic: 256, stage3_unit1_bn2[Float(256,28,28)], (Unnamed Layer* 192) [Constant]_output[Float(1,28,1)] -> stage3_unit1_relu1[Float(256,28,28)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn): stage3_unit1_conv1sc, Tactic: 5137655947464784826, _plus15[Float(128,28,28)] -> stage3_unit1_sc[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn): stage3_unit1_conv2 + _plus16, Tactic: 5137655947464784826, stage3_unit1_relu1[Float(256,28,28)], stage3_unit1_sc[Float(256,14,14)] -> _plus16[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit2_bn1 + stage3_unit2_bn1_scale, Tactic: 0, _plus16[Float(256,14,14)] -> stage3_unit2_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit2_conv1, Tactic: 1, stage3_unit2_bn1[Float(256,14,14)] -> stage3_unit2_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 206) [Constant], Tactic: 0, -> (Unnamed Layer* 206) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit2_relu1, Tactic: 128, stage3_unit2_bn2[Float(256,14,14)], (Unnamed Layer* 206) [Constant]_output[Float(1,14,1)] -> stage3_unit2_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit2_conv2 + _plus17, Tactic: 1, stage3_unit2_relu1[Float(256,14,14)], _plus16[Float(256,14,14)] -> _plus17[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit3_bn1 + stage3_unit3_bn1_scale, Tactic: 0, _plus17[Float(256,14,14)] -> stage3_unit3_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit3_conv1, Tactic: 1, stage3_unit3_bn1[Float(256,14,14)] -> stage3_unit3_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 217) [Constant], Tactic: 0, -> (Unnamed Layer* 217) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit3_relu1, Tactic: 128, stage3_unit3_bn2[Float(256,14,14)], (Unnamed Layer* 217) [Constant]_output[Float(1,14,1)] -> stage3_unit3_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit3_conv2 + _plus18, Tactic: 3827454225649558724, stage3_unit3_relu1[Float(256,14,14)], _plus17[Float(256,14,14)] -> _plus18[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit4_bn1 + stage3_unit4_bn1_scale, Tactic: 0, _plus18[Float(256,14,14)] -> stage3_unit4_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit4_conv1, Tactic: 1, stage3_unit4_bn1[Float(256,14,14)] -> stage3_unit4_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 228) [Constant], Tactic: 0, -> (Unnamed Layer* 228) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit4_relu1, Tactic: 256, stage3_unit4_bn2[Float(256,14,14)], (Unnamed Layer* 228) [Constant]_output[Float(1,14,1)] -> stage3_unit4_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit4_conv2 + _plus19, Tactic: 1, stage3_unit4_relu1[Float(256,14,14)], _plus18[Float(256,14,14)] -> _plus19[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit5_bn1 + stage3_unit5_bn1_scale, Tactic: 0, _plus19[Float(256,14,14)] -> stage3_unit5_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit5_conv1, Tactic: 3827454225649558724, stage3_unit5_bn1[Float(256,14,14)] -> stage3_unit5_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 239) [Constant], Tactic: 0, -> (Unnamed Layer* 239) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit5_relu1, Tactic: 128, stage3_unit5_bn2[Float(256,14,14)], (Unnamed Layer* 239) [Constant]_output[Float(1,14,1)] -> stage3_unit5_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit5_conv2 + _plus20, Tactic: 1, stage3_unit5_relu1[Float(256,14,14)], _plus19[Float(256,14,14)] -> _plus20[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit6_bn1 + stage3_unit6_bn1_scale, Tactic: 0, _plus20[Float(256,14,14)] -> stage3_unit6_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit6_conv1, Tactic: 3827454225649558724, stage3_unit6_bn1[Float(256,14,14)] -> stage3_unit6_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 250) [Constant], Tactic: 0, -> (Unnamed Layer* 250) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit6_relu1, Tactic: 128, stage3_unit6_bn2[Float(256,14,14)], (Unnamed Layer* 250) [Constant]_output[Float(1,14,1)] -> stage3_unit6_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit6_conv2 + _plus21, Tactic: 1, stage3_unit6_relu1[Float(256,14,14)], _plus20[Float(256,14,14)] -> _plus21[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit7_bn1 + stage3_unit7_bn1_scale, Tactic: 0, _plus21[Float(256,14,14)] -> stage3_unit7_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit7_conv1, Tactic: 1, stage3_unit7_bn1[Float(256,14,14)] -> stage3_unit7_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 261) [Constant], Tactic: 0, -> (Unnamed Layer* 261) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit7_relu1, Tactic: 256, stage3_unit7_bn2[Float(256,14,14)], (Unnamed Layer* 261) [Constant]_output[Float(1,14,1)] -> stage3_unit7_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit7_conv2 + _plus22, Tactic: 1, stage3_unit7_relu1[Float(256,14,14)], _plus21[Float(256,14,14)] -> _plus22[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit8_bn1 + stage3_unit8_bn1_scale, Tactic: 0, _plus22[Float(256,14,14)] -> stage3_unit8_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit8_conv1, Tactic: 3827454225649558724, stage3_unit8_bn1[Float(256,14,14)] -> stage3_unit8_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 272) [Constant], Tactic: 0, -> (Unnamed Layer* 272) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit8_relu1, Tactic: 256, stage3_unit8_bn2[Float(256,14,14)], (Unnamed Layer* 272) [Constant]_output[Float(1,14,1)] -> stage3_unit8_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit8_conv2 + _plus23, Tactic: 1, stage3_unit8_relu1[Float(256,14,14)], _plus22[Float(256,14,14)] -> _plus23[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit9_bn1 + stage3_unit9_bn1_scale, Tactic: 0, _plus23[Float(256,14,14)] -> stage3_unit9_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit9_conv1, Tactic: 3827454225649558724, stage3_unit9_bn1[Float(256,14,14)] -> stage3_unit9_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 283) [Constant], Tactic: 0, -> (Unnamed Layer* 283) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit9_relu1, Tactic: 128, stage3_unit9_bn2[Float(256,14,14)], (Unnamed Layer* 283) [Constant]_output[Float(1,14,1)] -> stage3_unit9_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit9_conv2 + _plus24, Tactic: 1, stage3_unit9_relu1[Float(256,14,14)], _plus23[Float(256,14,14)] -> _plus24[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit9_relu1, Tactic: 128, stage3_unit9_bn2[Float(256,14,14)], (Unnamed Layer* 283) [Constant]_output[Float(1,14,1)] -> stage3_unit9_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit9_conv2 + _plus24, Tactic: 1, stage3_unit9_relu1[Float(256,14,14)], _plus23[Float(256,14,14)] -> _plus24[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit10_bn1 + stage3_unit10_bn1_scale, Tactic: 0, _plus24[Float(256,14,14)] -> stage3_unit10_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit10_conv1, Tactic: 1, stage3_unit10_bn1[Float(256,14,14)] -> stage3_unit10_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 294) [Constant], Tactic: 0, -> (Unnamed Layer* 294) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit10_relu1, Tactic: 256, stage3_unit10_bn2[Float(256,14,14)], (Unnamed Layer* 294) [Constant]_output[Float(1,14,1)] -> stage3_unit10_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit10_conv2 + _plus25, Tactic: 3827454225649558724, stage3_unit10_relu1[Float(256,14,14)], _plus24[Float(256,14,14)] -> _plus25[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit11_bn1 + stage3_unit11_bn1_scale, Tactic: 0, _plus25[Float(256,14,14)] -> stage3_unit11_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit11_conv1, Tactic: 1, stage3_unit11_bn1[Float(256,14,14)] -> stage3_unit11_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 305) [Constant], Tactic: 0, -> (Unnamed Layer* 305) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit11_relu1, Tactic: 128, stage3_unit11_bn2[Float(256,14,14)], (Unnamed Layer* 305) [Constant]_output[Float(1,14,1)] -> stage3_unit11_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit11_conv2 + _plus26, Tactic: 3827454225649558724, stage3_unit11_relu1[Float(256,14,14)], _plus25[Float(256,14,14)] -> _plus26[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit12_bn1 + stage3_unit12_bn1_scale, Tactic: 0, _plus26[Float(256,14,14)] -> stage3_unit12_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit12_conv1, Tactic: 1, stage3_unit12_bn1[Float(256,14,14)] -> stage3_unit12_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 316) [Constant], Tactic: 0, -> (Unnamed Layer* 316) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit12_relu1, Tactic: 128, stage3_unit12_bn2[Float(256,14,14)], (Unnamed Layer* 316) [Constant]_output[Float(1,14,1)] -> stage3_unit12_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit12_conv2 + _plus27, Tactic: 1, stage3_unit12_relu1[Float(256,14,14)], _plus26[Float(256,14,14)] -> _plus27[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit13_bn1 + stage3_unit13_bn1_scale, Tactic: 0, _plus27[Float(256,14,14)] -> stage3_unit13_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit13_conv1, Tactic: 1, stage3_unit13_bn1[Float(256,14,14)] -> stage3_unit13_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 327) [Constant], Tactic: 0, -> (Unnamed Layer* 327) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit13_relu1, Tactic: 128, stage3_unit13_bn2[Float(256,14,14)], (Unnamed Layer* 327) [Constant]_output[Float(1,14,1)] -> stage3_unit13_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit13_conv2 + _plus28, Tactic: 1, stage3_unit13_relu1[Float(256,14,14)], _plus27[Float(256,14,14)] -> _plus28[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit14_bn1 + stage3_unit14_bn1_scale, Tactic: 0, _plus28[Float(256,14,14)] -> stage3_unit14_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit14_conv1, Tactic: 1, stage3_unit14_bn1[Float(256,14,14)] -> stage3_unit14_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 338) [Constant], Tactic: 0, -> (Unnamed Layer* 338) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit14_relu1, Tactic: 128, stage3_unit14_bn2[Float(256,14,14)], (Unnamed Layer* 338) [Constant]_output[Float(1,14,1)] -> stage3_unit14_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit14_conv2 + _plus29, Tactic: 3827454225649558724, stage3_unit14_relu1[Float(256,14,14)], _plus28[Float(256,14,14)] -> _plus29[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit15_bn1 + stage3_unit15_bn1_scale, Tactic: 0, _plus29[Float(256,14,14)] -> stage3_unit15_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit15_conv1, Tactic: 1, stage3_unit15_bn1[Float(256,14,14)] -> stage3_unit15_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 349) [Constant], Tactic: 0, -> (Unnamed Layer* 349) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit15_relu1, Tactic: 256, stage3_unit15_bn2[Float(256,14,14)], (Unnamed Layer* 349) [Constant]_output[Float(1,14,1)] -> stage3_unit15_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit15_conv2 + _plus30, Tactic: 3827454225649558724, stage3_unit15_relu1[Float(256,14,14)], _plus29[Float(256,14,14)] -> _plus30[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit16_bn1 + stage3_unit16_bn1_scale, Tactic: 0, _plus30[Float(256,14,14)] -> stage3_unit16_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit16_conv1, Tactic: 1, stage3_unit16_bn1[Float(256,14,14)] -> stage3_unit16_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 360) [Constant], Tactic: 0, -> (Unnamed Layer* 360) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit16_relu1, Tactic: 256, stage3_unit16_bn2[Float(256,14,14)], (Unnamed Layer* 360) [Constant]_output[Float(1,14,1)] -> stage3_unit16_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit16_conv2 + _plus31, Tactic: 3827454225649558724, stage3_unit16_relu1[Float(256,14,14)], _plus30[Float(256,14,14)] -> _plus31[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit17_bn1 + stage3_unit17_bn1_scale, Tactic: 0, _plus31[Float(256,14,14)] -> stage3_unit17_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit17_conv1, Tactic: 3827454225649558724, stage3_unit17_bn1[Float(256,14,14)] -> stage3_unit17_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 371) [Constant], Tactic: 0, -> (Unnamed Layer* 371) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit17_relu1, Tactic: 256, stage3_unit17_bn2[Float(256,14,14)], (Unnamed Layer* 371) [Constant]_output[Float(1,14,1)] -> stage3_unit17_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit17_conv2 + _plus32, Tactic: 1, stage3_unit17_relu1[Float(256,14,14)], _plus31[Float(256,14,14)] -> _plus32[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit18_bn1 + stage3_unit18_bn1_scale, Tactic: 0, _plus32[Float(256,14,14)] -> stage3_unit18_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit18_conv1, Tactic: 3827454225649558724, stage3_unit18_bn1[Float(256,14,14)] -> stage3_unit18_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 382) [Constant], Tactic: 0, -> (Unnamed Layer* 382) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit18_relu1, Tactic: 128, stage3_unit18_bn2[Float(256,14,14)], (Unnamed Layer* 382) [Constant]_output[Float(1,14,1)] -> stage3_unit18_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit18_conv2 + _plus33, Tactic: 3827454225649558724, stage3_unit18_relu1[Float(256,14,14)], _plus32[Float(256,14,14)] -> _plus33[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit19_bn1 + stage3_unit19_bn1_scale, Tactic: 0, _plus33[Float(256,14,14)] -> stage3_unit19_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit19_conv1, Tactic: 3827454225649558724, stage3_unit19_bn1[Float(256,14,14)] -> stage3_unit19_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 393) [Constant], Tactic: 0, -> (Unnamed Layer* 393) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit19_relu1, Tactic: 256, stage3_unit19_bn2[Float(256,14,14)], (Unnamed Layer* 393) [Constant]_output[Float(1,14,1)] -> stage3_unit19_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit19_conv2 + _plus34, Tactic: 3827454225649558724, stage3_unit19_relu1[Float(256,14,14)], _plus33[Float(256,14,14)] -> _plus34[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit19_relu1, Tactic: 256, stage3_unit19_bn2[Float(256,14,14)], (Unnamed Layer* 393) [Constant]_output[Float(1,14,1)] -> stage3_unit19_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit19_conv2 + _plus34, Tactic: 3827454225649558724, stage3_unit19_relu1[Float(256,14,14)], _plus33[Float(256,14,14)] -> _plus34[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit20_bn1 + stage3_unit20_bn1_scale, Tactic: 0, _plus34[Float(256,14,14)] -> stage3_unit20_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit20_conv1, Tactic: 1, stage3_unit20_bn1[Float(256,14,14)] -> stage3_unit20_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 404) [Constant], Tactic: 0, -> (Unnamed Layer* 404) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit20_relu1, Tactic: 256, stage3_unit20_bn2[Float(256,14,14)], (Unnamed Layer* 404) [Constant]_output[Float(1,14,1)] -> stage3_unit20_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit20_conv2 + _plus35, Tactic: 3827454225649558724, stage3_unit20_relu1[Float(256,14,14)], _plus34[Float(256,14,14)] -> _plus35[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit21_bn1 + stage3_unit21_bn1_scale, Tactic: 0, _plus35[Float(256,14,14)] -> stage3_unit21_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit21_conv1, Tactic: 1, stage3_unit21_bn1[Float(256,14,14)] -> stage3_unit21_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 415) [Constant], Tactic: 0, -> (Unnamed Layer* 415) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit21_relu1, Tactic: 128, stage3_unit21_bn2[Float(256,14,14)], (Unnamed Layer* 415) [Constant]_output[Float(1,14,1)] -> stage3_unit21_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit21_conv2 + _plus36, Tactic: 1, stage3_unit21_relu1[Float(256,14,14)], _plus35[Float(256,14,14)] -> _plus36[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit22_bn1 + stage3_unit22_bn1_scale, Tactic: 0, _plus36[Float(256,14,14)] -> stage3_unit22_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit22_conv1, Tactic: 1, stage3_unit22_bn1[Float(256,14,14)] -> stage3_unit22_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 426) [Constant], Tactic: 0, -> (Unnamed Layer* 426) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit22_relu1, Tactic: 256, stage3_unit22_bn2[Float(256,14,14)], (Unnamed Layer* 426) [Constant]_output[Float(1,14,1)] -> stage3_unit22_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit22_conv2 + _plus37, Tactic: 1, stage3_unit22_relu1[Float(256,14,14)], _plus36[Float(256,14,14)] -> _plus37[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit23_bn1 + stage3_unit23_bn1_scale, Tactic: 0, _plus37[Float(256,14,14)] -> stage3_unit23_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit23_conv1, Tactic: 1, stage3_unit23_bn1[Float(256,14,14)] -> stage3_unit23_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 437) [Constant], Tactic: 0, -> (Unnamed Layer* 437) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit23_relu1, Tactic: 128, stage3_unit23_bn2[Float(256,14,14)], (Unnamed Layer* 437) [Constant]_output[Float(1,14,1)] -> stage3_unit23_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit23_conv2 + _plus38, Tactic: 1, stage3_unit23_relu1[Float(256,14,14)], _plus37[Float(256,14,14)] -> _plus38[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit24_bn1 + stage3_unit24_bn1_scale, Tactic: 0, _plus38[Float(256,14,14)] -> stage3_unit24_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit24_conv1, Tactic: 1, stage3_unit24_bn1[Float(256,14,14)] -> stage3_unit24_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 448) [Constant], Tactic: 0, -> (Unnamed Layer* 448) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit24_relu1, Tactic: 256, stage3_unit24_bn2[Float(256,14,14)], (Unnamed Layer* 448) [Constant]_output[Float(1,14,1)] -> stage3_unit24_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit24_conv2 + _plus39, Tactic: 3827454225649558724, stage3_unit24_relu1[Float(256,14,14)], _plus38[Float(256,14,14)] -> _plus39[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit25_bn1 + stage3_unit25_bn1_scale, Tactic: 0, _plus39[Float(256,14,14)] -> stage3_unit25_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit25_conv1, Tactic: 1, stage3_unit25_bn1[Float(256,14,14)] -> stage3_unit25_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 459) [Constant], Tactic: 0, -> (Unnamed Layer* 459) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit25_relu1, Tactic: 128, stage3_unit25_bn2[Float(256,14,14)], (Unnamed Layer* 459) [Constant]_output[Float(1,14,1)] -> stage3_unit25_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit25_conv2 + _plus40, Tactic: 3827454225649558724, stage3_unit25_relu1[Float(256,14,14)], _plus39[Float(256,14,14)] -> _plus40[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit26_bn1 + stage3_unit26_bn1_scale, Tactic: 0, _plus40[Float(256,14,14)] -> stage3_unit26_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit26_conv1, Tactic: 1, stage3_unit26_bn1[Float(256,14,14)] -> stage3_unit26_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 470) [Constant], Tactic: 0, -> (Unnamed Layer* 470) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit26_relu1, Tactic: 128, stage3_unit26_bn2[Float(256,14,14)], (Unnamed Layer* 470) [Constant]_output[Float(1,14,1)] -> stage3_unit26_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit26_conv2 + _plus41, Tactic: 3827454225649558724, stage3_unit26_relu1[Float(256,14,14)], _plus40[Float(256,14,14)] -> _plus41[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit27_bn1 + stage3_unit27_bn1_scale, Tactic: 0, _plus41[Float(256,14,14)] -> stage3_unit27_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit27_conv1, Tactic: 3827454225649558724, stage3_unit27_bn1[Float(256,14,14)] -> stage3_unit27_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 481) [Constant], Tactic: 0, -> (Unnamed Layer* 481) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit27_relu1, Tactic: 256, stage3_unit27_bn2[Float(256,14,14)], (Unnamed Layer* 481) [Constant]_output[Float(1,14,1)] -> stage3_unit27_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit27_conv2 + _plus42, Tactic: 3827454225649558724, stage3_unit27_relu1[Float(256,14,14)], _plus41[Float(256,14,14)] -> _plus42[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit28_bn1 + stage3_unit28_bn1_scale, Tactic: 0, _plus42[Float(256,14,14)] -> stage3_unit28_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit28_conv1, Tactic: 1, stage3_unit28_bn1[Float(256,14,14)] -> stage3_unit28_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 492) [Constant], Tactic: 0, -> (Unnamed Layer* 492) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit28_relu1, Tactic: 256, stage3_unit28_bn2[Float(256,14,14)], (Unnamed Layer* 492) [Constant]_output[Float(1,14,1)] -> stage3_unit28_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit28_conv2 + _plus43, Tactic: 3827454225649558724, stage3_unit28_relu1[Float(256,14,14)], _plus42[Float(256,14,14)] -> _plus43[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit29_bn1 + stage3_unit29_bn1_scale, Tactic: 0, _plus43[Float(256,14,14)] -> stage3_unit29_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit29_conv1, Tactic: 3827454225649558724, stage3_unit29_bn1[Float(256,14,14)] -> stage3_unit29_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 503) [Constant], Tactic: 0, -> (Unnamed Layer* 503) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit29_relu1, Tactic: 256, stage3_unit29_bn2[Float(256,14,14)], (Unnamed Layer* 503) [Constant]_output[Float(1,14,1)] -> stage3_unit29_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit29_conv2 + _plus44, Tactic: 3827454225649558724, stage3_unit29_relu1[Float(256,14,14)], _plus43[Float(256,14,14)] -> _plus44[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit29_relu1, Tactic: 256, stage3_unit29_bn2[Float(256,14,14)], (Unnamed Layer* 503) [Constant]_output[Float(1,14,1)] -> stage3_unit29_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit29_conv2 + _plus44, Tactic: 3827454225649558724, stage3_unit29_relu1[Float(256,14,14)], _plus43[Float(256,14,14)] -> _plus44[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage3_unit30_bn1 + stage3_unit30_bn1_scale, Tactic: 0, _plus44[Float(256,14,14)] -> stage3_unit30_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage3_unit30_conv1, Tactic: 1, stage3_unit30_bn1[Float(256,14,14)] -> stage3_unit30_bn2[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 514) [Constant], Tactic: 0, -> (Unnamed Layer* 514) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage3_unit30_relu1, Tactic: 128, stage3_unit30_bn2[Float(256,14,14)], (Unnamed Layer* 514) [Constant]_output[Float(1,14,1)] -> stage3_unit30_relu1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage3_unit30_conv2 + _plus45, Tactic: 3827454225649558724, stage3_unit30_relu1[Float(256,14,14)], _plus44[Float(256,14,14)] -> _plus45[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage4_unit1_bn1 + stage4_unit1_bn1_scale, Tactic: 0, _plus45[Float(256,14,14)] -> stage4_unit1_bn1[Float(256,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn_winograd): stage4_unit1_conv1, Tactic: 3827454225649558724, stage4_unit1_bn1[Float(256,14,14)] -> stage4_unit1_bn2[Float(512,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 525) [Constant], Tactic: 0, -> (Unnamed Layer* 525) [Constant]_output[Float(1,14,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage4_unit1_relu1, Tactic: 128, stage4_unit1_bn2[Float(512,14,14)], (Unnamed Layer* 525) [Constant]_output[Float(1,14,1)] -> stage4_unit1_relu1[Float(512,14,14)]
[03/25/2020-17:14:04] [V] [TRT] Layer(scudnn): stage4_unit1_conv1sc, Tactic: 5137655947464784826, _plus45[Float(256,14,14)] -> stage4_unit1_sc[Float(512,7,7)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Convolution): stage4_unit1_conv2 + _plus46, Tactic: 2, stage4_unit1_relu1[Float(512,14,14)], stage4_unit1_sc[Float(512,7,7)] -> _plus46[Float(512,7,7)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage4_unit2_bn1 + stage4_unit2_bn1_scale, Tactic: 0, _plus46[Float(512,7,7)] -> stage4_unit2_bn1[Float(512,7,7)]
[03/25/2020-17:14:04] [V] [TRT] Layer(FusedConvActDirect): stage4_unit2_conv1, Tactic: 134, stage4_unit2_bn1[Float(512,7,7)] -> stage4_unit2_bn2[Float(512,7,7)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 539) [Constant], Tactic: 0, -> (Unnamed Layer* 539) [Constant]_output[Float(1,7,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage4_unit2_relu1, Tactic: 128, stage4_unit2_bn2[Float(512,7,7)], (Unnamed Layer* 539) [Constant]_output[Float(1,7,1)] -> stage4_unit2_relu1[Float(512,7,7)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage4_unit2_conv2 + _plus47, Tactic: 1, stage4_unit2_relu1[Float(512,7,7)], _plus46[Float(512,7,7)] -> _plus47[Float(512,7,7)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): stage4_unit3_bn1 + stage4_unit3_bn1_scale, Tactic: 0, _plus47[Float(512,7,7)] -> stage4_unit3_bn1[Float(512,7,7)]
[03/25/2020-17:14:04] [V] [TRT] Layer(FusedConvActDirect): stage4_unit3_conv1, Tactic: 134, stage4_unit3_bn1[Float(512,7,7)] -> stage4_unit3_bn2[Float(512,7,7)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Constant): (Unnamed Layer* 550) [Constant], Tactic: 0, -> (Unnamed Layer* 550) [Constant]_output[Float(1,7,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(PointWise): stage4_unit3_relu1, Tactic: 128, stage4_unit3_bn2[Float(512,7,7)], (Unnamed Layer* 550) [Constant]_output[Float(1,7,1)] -> stage4_unit3_relu1[Float(512,7,7)]
[03/25/2020-17:14:04] [V] [TRT] Layer(CustomWinograd): stage4_unit3_conv2 + _plus48, Tactic: 1, stage4_unit3_relu1[Float(512,7,7)], _plus47[Float(512,7,7)] -> _plus48[Float(512,7,7)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): bn1 + bn1_scale, Tactic: 0, _plus48[Float(512,7,7)] -> dropout0[Float(512,7,7)]
[03/25/2020-17:14:04] [V] [TRT] Layer(FullyConnected): pre_fc1, Tactic: 1, dropout0[Float(512,7,7)] -> pre_fc1[Float(512,1,1)]
[03/25/2020-17:14:04] [V] [TRT] Layer(Scale): fc1 + fc1_scale, Tactic: 0, pre_fc1[Float(512,1,1)] -> fc1[Float(512,1,1)]
[03/25/2020-17:14:04] [03/25/2020-17:14:08] [I] Warmup completed 1 queries over 200 ms
[03/25/2020-17:14:08] [I] Timing trace has 587 queries over 3.00996 s
[03/25/2020-17:14:08] [I] Trace averages of 10 runs:
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 6.13028 ms - Host latency: 6.16138 ms (end to end 12.0118 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.89026 ms - Host latency: 5.92362 ms (end to end 11.2615 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.71106 ms - Host latency: 5.74456 ms (end to end 11.2214 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.70595 ms - Host latency: 5.73772 ms (end to end 10.584 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.43181 ms - Host latency: 5.4646 ms (end to end 10.5996 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.24144 ms - Host latency: 5.27394 ms (end to end 10.2759 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.222 ms - Host latency: 5.25414 ms (end to end 10.2496 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.14816 ms - Host latency: 5.18033 ms (end to end 9.61559 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.02264 ms - Host latency: 5.05484 ms (end to end 9.83767 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.02344 ms - Host latency: 5.05515 ms (end to end 9.84534 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.02608 ms - Host latency: 5.05745 ms (end to end 9.85288 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.0425 ms - Host latency: 5.07274 ms (end to end 9.20037 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.0519 ms - Host latency: 5.0841 ms (end to end 9.89714 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.05222 ms - Host latency: 5.08479 ms (end to end 9.90218 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04842 ms - Host latency: 5.07981 ms (end to end 9.89466 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04648 ms - Host latency: 5.0785 ms (end to end 9.89768 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04832 ms - Host latency: 5.08062 ms (end to end 9.89811 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.05039 ms - Host latency: 5.08474 ms (end to end 9.91854 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.05048 ms - Host latency: 5.08179 ms (end to end 9.99187 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.0512 ms - Host latency: 5.08237 ms (end to end 9.91799 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04685 ms - Host latency: 5.07816 ms (end to end 9.91488 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04823 ms - Host latency: 5.07889 ms (end to end 9.9121 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04474 ms - Host latency: 5.07551 ms (end to end 9.9084 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04823 ms - Host latency: 5.07889 ms (end to end 9.9121 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04474 ms - Host latency: 5.07551 ms (end to end 9.9084 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04586 ms - Host latency: 5.07714 ms (end to end 9.9125 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04904 ms - Host latency: 5.07987 ms (end to end 9.9126 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04705 ms - Host latency: 5.07745 ms (end to end 9.90737 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04814 ms - Host latency: 5.07946 ms (end to end 9.91608 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04509 ms - Host latency: 5.07633 ms (end to end 9.91143 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.0468 ms - Host latency: 5.07793 ms (end to end 9.91688 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04841 ms - Host latency: 5.07917 ms (end to end 9.91551 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04529 ms - Host latency: 5.07672 ms (end to end 9.90726 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.0467 ms - Host latency: 5.07747 ms (end to end 9.91728 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04617 ms - Host latency: 5.07726 ms (end to end 9.90934 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.05204 ms - Host latency: 5.08281 ms (end to end 9.91638 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04856 ms - Host latency: 5.07975 ms (end to end 9.91348 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04738 ms - Host latency: 5.07809 ms (end to end 9.91643 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04827 ms - Host latency: 5.07969 ms (end to end 9.91326 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04692 ms - Host latency: 5.07737 ms (end to end 9.90974 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04641 ms - Host latency: 5.07761 ms (end to end 9.91387 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.0491 ms - Host latency: 5.07952 ms (end to end 9.91646 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04622 ms - Host latency: 5.08196 ms (end to end 9.9386 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.05193 ms - Host latency: 5.08367 ms (end to end 9.94683 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04937 ms - Host latency: 5.08359 ms (end to end 9.95549 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.05305 ms - Host latency: 5.0855 ms (end to end 9.95464 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.051 ms - Host latency: 5.08147 ms (end to end 9.94031 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04595 ms - Host latency: 5.07942 ms (end to end 9.9156 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04983 ms - Host latency: 5.08032 ms (end to end 9.91716 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04668 ms - Host latency: 5.07778 ms (end to end 9.90874 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04858 ms - Host latency: 5.08 ms (end to end 9.91248 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04958 ms - Host latency: 5.07969 ms (end to end 9.91143 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04502 ms - Host latency: 5.07627 ms (end to end 9.90671 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04485 ms - Host latency: 5.07593 ms (end to end 9.90537 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04771 ms - Host latency: 5.07876 ms (end to end 9.91702 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04717 ms - Host latency: 5.07881 ms (end to end 9.91228 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04299 ms - Host latency: 5.07373 ms (end to end 9.90508 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04873 ms - Host latency: 5.08101 ms (end to end 9.90645 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.05 ms - Host latency: 5.08298 ms (end to end 9.8959 ms)
[03/25/2020-17:14:08] [I] Average on 10 runs - GPU latency: 5.04712 ms - Host latency: 5.07808 ms (end to end 9.89204 ms)
[03/25/2020-17:14:08] [I] Host latency
[03/25/2020-17:14:08] [I] min: 5.04639 ms (end to end 5.33325 ms)
[03/25/2020-17:14:08] [I] max: 6.17398 ms (end to end 12.0827 ms)
[03/25/2020-17:14:08] [I] mean: 5.14821 ms (end to end 10.0103 ms)
[03/25/2020-17:14:08] [I] median: 5.08014 ms (end to end 9.91394 ms)
[03/25/2020-17:14:08] [I] percentile: 6.16275 ms at 99% (end to end 12.0636 ms at 99%)
[03/25/2020-17:14:08] [I] throughput: 195.019 qps
[03/25/2020-17:14:08] [I] walltime: 3.00996 s
[03/25/2020-17:14:08] [I] GPU Compute
[03/25/2020-17:14:08] [I] min: 5.01349 ms
[03/25/2020-17:14:08] [I] max: 6.14093 ms
[03/25/2020-17:14:08] [I] mean: 5.11657 ms
[03/25/2020-17:14:08] [I] median: 5.04932 ms
[03/25/2020-17:14:08] [I] percentile: 6.1317 ms at 99%
[03/25/2020-17:14:08] [I] total compute time: 3.00343 s
[03/25/2020-17:14:08] [I] percentile: 6.1317 ms at 99%
[03/25/2020-17:14:08] [I] total compute time: 3.00343 s
&&&& PASSED TensorRT.trtexec # ./trtexec --deploy=models/resnet101.prototxt --output=fc1 --verbose

Hi, any news regarding this issue?

Yes, I met the same issue. Any reply?