$ /usr/src/tensorrt/bin/trtexec --onnx=reduced.onnx --shapes=input:1x3x480x640 --workspace=1000 --avgRuns=10 --verbose
&&&& RUNNING TensorRT.trtexec # /usr/src/tensorrt/bin/trtexec --onnx=reduced.onnx --shapes=input:1x3x480x640 --workspace=1000 --avgRuns=10 --verbose
[11/17/2022-03:14:13] [I] === Model Options ===
[11/17/2022-03:14:13] [I] Format: ONNX
[11/17/2022-03:14:13] [I] Model: reduced.onnx
[11/17/2022-03:14:13] [I] Output:
[11/17/2022-03:14:13] [I] === Build Options ===
[11/17/2022-03:14:13] [I] Max batch: explicit
[11/17/2022-03:14:13] [I] Workspace: 1000 MB
[11/17/2022-03:14:13] [I] minTiming: 1
[11/17/2022-03:14:13] [I] avgTiming: 8
[11/17/2022-03:14:13] [I] Precision: FP32
[11/17/2022-03:14:13] [I] Calibration: 
[11/17/2022-03:14:13] [I] Safe mode: Disabled
[11/17/2022-03:14:13] [I] Save engine: 
[11/17/2022-03:14:13] [I] Load engine: 
[11/17/2022-03:14:13] [I] Builder Cache: Enabled
[11/17/2022-03:14:13] [I] NVTX verbosity: 0
[11/17/2022-03:14:13] [I] Inputs format: fp32:CHW
[11/17/2022-03:14:13] [I] Outputs format: fp32:CHW
[11/17/2022-03:14:13] [I] Input build shape: input=1x3x480x640+1x3x480x640+1x3x480x640
[11/17/2022-03:14:13] [I] Input calibration shapes: model
[11/17/2022-03:14:13] [I] === System Options ===
[11/17/2022-03:14:13] [I] Device: 0
[11/17/2022-03:14:13] [I] DLACore: 
[11/17/2022-03:14:13] [I] Plugins:
[11/17/2022-03:14:13] [I] === Inference Options ===
[11/17/2022-03:14:13] [I] Batch: Explicit
[11/17/2022-03:14:13] [I] Input inference shape: input=1x3x480x640
[11/17/2022-03:14:13] [I] Iterations: 10
[11/17/2022-03:14:13] [I] Duration: 3s (+ 200ms warm up)
[11/17/2022-03:14:13] [I] Sleep time: 0ms
[11/17/2022-03:14:13] [I] Streams: 1
[11/17/2022-03:14:13] [I] ExposeDMA: Disabled
[11/17/2022-03:14:13] [I] Spin-wait: Disabled
[11/17/2022-03:14:13] [I] Multithreading: Disabled
[11/17/2022-03:14:13] [I] CUDA Graph: Disabled
[11/17/2022-03:14:13] [I] Skip inference: Disabled
[11/17/2022-03:14:13] [I] Inputs:
[11/17/2022-03:14:13] [I] === Reporting Options ===
[11/17/2022-03:14:13] [I] Verbose: Enabled
[11/17/2022-03:14:13] [I] Averages: 10 inferences
[11/17/2022-03:14:13] [I] Percentile: 99
[11/17/2022-03:14:13] [I] Dump output: Disabled
[11/17/2022-03:14:13] [I] Profile: Disabled
[11/17/2022-03:14:13] [I] Export timing to JSON file: 
[11/17/2022-03:14:13] [I] Export output to JSON file: 
[11/17/2022-03:14:13] [I] Export profile to JSON file: 
[11/17/2022-03:14:13] [I] 
[11/17/2022-03:14:13] [V] [TRT] Registered plugin creator - ::GridAnchor_TRT version 1
[11/17/2022-03:14:13] [V] [TRT] Registered plugin creator - ::NMS_TRT version 1
[11/17/2022-03:14:13] [V] [TRT] Registered plugin creator - ::Reorg_TRT version 1
[11/17/2022-03:14:13] [V] [TRT] Registered plugin creator - ::Region_TRT version 1
[11/17/2022-03:14:13] [V] [TRT] Registered plugin creator - ::Clip_TRT version 1
[11/17/2022-03:14:13] [V] [TRT] Registered plugin creator - ::LReLU_TRT version 1
[11/17/2022-03:14:13] [V] [TRT] Registered plugin creator - ::PriorBox_TRT version 1
[11/17/2022-03:14:13] [V] [TRT] Registered plugin creator - ::Normalize_TRT version 1
[11/17/2022-03:14:13] [V] [TRT] Registered plugin creator - ::RPROI_TRT version 1
[11/17/2022-03:14:13] [V] [TRT] Registered plugin creator - ::BatchedNMS_TRT version 1
[11/17/2022-03:14:13] [V] [TRT] Registered plugin creator - ::FlattenConcat_TRT version 1
[11/17/2022-03:14:13] [V] [TRT] Registered plugin creator - ::CropAndResize version 1
[11/17/2022-03:14:13] [V] [TRT] Registered plugin creator - ::DetectionLayer_TRT version 1
[11/17/2022-03:14:13] [V] [TRT] Registered plugin creator - ::Proposal version 1
[11/17/2022-03:14:13] [V] [TRT] Registered plugin creator - ::ProposalLayer_TRT version 1
[11/17/2022-03:14:13] [V] [TRT] Registered plugin creator - ::PyramidROIAlign_TRT version 1
[11/17/2022-03:14:13] [V] [TRT] Registered plugin creator - ::ResizeNearest_TRT version 1
[11/17/2022-03:14:13] [V] [TRT] Registered plugin creator - ::Split version 1
[11/17/2022-03:14:13] [V] [TRT] Registered plugin creator - ::SpecialSlice_TRT version 1
[11/17/2022-03:14:13] [V] [TRT] Registered plugin creator - ::InstanceNormalization_TRT version 1
----------------------------------------------------------------
Input filename:   reduced.onnx
ONNX IR version:  0.0.7
Opset version:    11
Producer name:    
Producer version: 
Domain:           
Model version:    0
Doc string:       
----------------------------------------------------------------
[11/17/2022-03:14:15] [V] [TRT] Plugin creator already registered - ::GridAnchor_TRT version 1
[11/17/2022-03:14:15] [V] [TRT] Plugin creator already registered - ::NMS_TRT version 1
[11/17/2022-03:14:15] [V] [TRT] Plugin creator already registered - ::Reorg_TRT version 1
[11/17/2022-03:14:15] [V] [TRT] Plugin creator already registered - ::Region_TRT version 1
[11/17/2022-03:14:15] [V] [TRT] Plugin creator already registered - ::Clip_TRT version 1
[11/17/2022-03:14:15] [V] [TRT] Plugin creator already registered - ::LReLU_TRT version 1
[11/17/2022-03:14:15] [V] [TRT] Plugin creator already registered - ::PriorBox_TRT version 1
[11/17/2022-03:14:15] [V] [TRT] Plugin creator already registered - ::Normalize_TRT version 1
[11/17/2022-03:14:15] [V] [TRT] Plugin creator already registered - ::RPROI_TRT version 1
[11/17/2022-03:14:15] [V] [TRT] Plugin creator already registered - ::BatchedNMS_TRT version 1
[11/17/2022-03:14:15] [V] [TRT] Plugin creator already registered - ::FlattenConcat_TRT version 1
[11/17/2022-03:14:15] [V] [TRT] Plugin creator already registered - ::CropAndResize version 1
[11/17/2022-03:14:15] [V] [TRT] Plugin creator already registered - ::DetectionLayer_TRT version 1
[11/17/2022-03:14:15] [V] [TRT] Plugin creator already registered - ::Proposal version 1
[11/17/2022-03:14:15] [V] [TRT] Plugin creator already registered - ::ProposalLayer_TRT version 1
[11/17/2022-03:14:15] [V] [TRT] Plugin creator already registered - ::PyramidROIAlign_TRT version 1
[11/17/2022-03:14:15] [V] [TRT] Plugin creator already registered - ::ResizeNearest_TRT version 1
[11/17/2022-03:14:15] [V] [TRT] Plugin creator already registered - ::Split version 1
[11/17/2022-03:14:15] [V] [TRT] Plugin creator already registered - ::SpecialSlice_TRT version 1
[11/17/2022-03:14:15] [V] [TRT] Plugin creator already registered - ::InstanceNormalization_TRT version 1
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:202: Adding network input: model_2/model/Conv1_relu/Relu6:0 with dtype: float32, dimensions: (1, 32, 240, 320)
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:116: Registering tensor: model_2/model/Conv1_relu/Relu6:0 for ONNX tensor: model_2/model/Conv1_relu/Relu6:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:90: Importing initializer: model_2/model/expanded_conv_depthwise/depthwise_weights_fused_bn
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:90: Importing initializer: model_2/model/expanded_conv_depthwise/depthwise_bias_fused_bn
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:90: Importing initializer: model_2/model/block_9_depthwise_relu/Relu6_min__196
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:90: Importing initializer: model_2/model/block_4_expand_relu/Relu6_max__92
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:90: Importing initializer: model_2/model/expanded_conv_project/Conv2D_weights_fused_bn
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:90: Importing initializer: model_2/model/expanded_conv_project/Conv2D_bias_fused_bn
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:90: Importing initializer: 317
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:90: Importing initializer: 319
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:90: Importing initializer: model_2/model/block_1_depthwise/depthwise_weights_fused_bn
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:90: Importing initializer: model_2/model/block_1_depthwise/depthwise_bias_fused_bn
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:90: Importing initializer: model_2/model/block_1_project/Conv2D_weights_fused_bn
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:90: Importing initializer: model_2/model/block_1_project/Conv2D_bias_fused_bn
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:90: Importing initializer: 321
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:90: Importing initializer: 323
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:90: Importing initializer: model_2/model/block_2_depthwise/depthwise_weights_fused_bn
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:90: Importing initializer: model_2/model/block_2_depthwise/depthwise_bias_fused_bn
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:90: Importing initializer: model_2/model/block_2_project/Conv2D_weights_fused_bn
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:90: Importing initializer: model_2/model/block_2_project/Conv2D_bias_fused_bn
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:103: Parsing node: model_2/model/expanded_conv_depthwise/depthwise [Conv]
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/Conv1_relu/Relu6:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/expanded_conv_depthwise/depthwise_weights_fused_bn
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/expanded_conv_depthwise/depthwise_bias_fused_bn
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:125: model_2/model/expanded_conv_depthwise/depthwise [Conv] inputs: [model_2/model/Conv1_relu/Relu6:0 -> (1, 32, 240, 320)], [model_2/model/expanded_conv_depthwise/depthwise_weights_fused_bn -> (32, 1, 3, 3)], [model_2/model/expanded_conv_depthwise/depthwise_bias_fused_bn -> (32)], 
[11/17/2022-03:14:15] [V] [TRT] builtin_op_importers.cpp:450: Convolution input dimensions: (1, 32, 240, 320)
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:141: Registering layer: model_2/model/expanded_conv_depthwise/depthwise for ONNX node: model_2/model/expanded_conv_depthwise/depthwise
[11/17/2022-03:14:15] [V] [TRT] builtin_op_importers.cpp:533: Using kernel: (3, 3), strides: (1, 1), prepadding: (1, 1), postpadding: (1, 1), dilations: (1, 1), numOutputs: 32
[11/17/2022-03:14:15] [V] [TRT] builtin_op_importers.cpp:534: Convolution output dimensions: (1, 32, 240, 320)
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:116: Registering tensor: model_2/model/expanded_conv_depthwise_BN/FusedBatchNormV3:0 for ONNX tensor: model_2/model/expanded_conv_depthwise_BN/FusedBatchNormV3:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:179: model_2/model/expanded_conv_depthwise/depthwise [Conv] outputs: [model_2/model/expanded_conv_depthwise_BN/FusedBatchNormV3:0 -> (1, 32, 240, 320)], 
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:103: Parsing node: model_2/model/expanded_conv_depthwise_relu/Relu6 [Clip]
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/expanded_conv_depthwise_BN/FusedBatchNormV3:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_9_depthwise_relu/Relu6_min__196
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_4_expand_relu/Relu6_max__92
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:125: model_2/model/expanded_conv_depthwise_relu/Relu6 [Clip] inputs: [model_2/model/expanded_conv_depthwise_BN/FusedBatchNormV3:0 -> (1, 32, 240, 320)], [model_2/model/block_9_depthwise_relu/Relu6_min__196 -> ()], [model_2/model/block_4_expand_relu/Relu6_max__92 -> ()], 
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:141: Registering layer: model_2/model/expanded_conv_depthwise_relu/Relu6 for ONNX node: model_2/model/expanded_conv_depthwise_relu/Relu6
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:116: Registering tensor: model_2/model/expanded_conv_depthwise_relu/Relu6:0 for ONNX tensor: model_2/model/expanded_conv_depthwise_relu/Relu6:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:179: model_2/model/expanded_conv_depthwise_relu/Relu6 [Clip] outputs: [model_2/model/expanded_conv_depthwise_relu/Relu6:0 -> (1, 32, 240, 320)], 
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:103: Parsing node: model_2/model/expanded_conv_project/Conv2D [Conv]
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/expanded_conv_depthwise_relu/Relu6:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/expanded_conv_project/Conv2D_weights_fused_bn
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/expanded_conv_project/Conv2D_bias_fused_bn
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:125: model_2/model/expanded_conv_project/Conv2D [Conv] inputs: [model_2/model/expanded_conv_depthwise_relu/Relu6:0 -> (1, 32, 240, 320)], [model_2/model/expanded_conv_project/Conv2D_weights_fused_bn -> (16, 32, 1, 1)], [model_2/model/expanded_conv_project/Conv2D_bias_fused_bn -> (16)], 
[11/17/2022-03:14:15] [V] [TRT] builtin_op_importers.cpp:450: Convolution input dimensions: (1, 32, 240, 320)
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:141: Registering layer: model_2/model/expanded_conv_project/Conv2D for ONNX node: model_2/model/expanded_conv_project/Conv2D
[11/17/2022-03:14:15] [V] [TRT] builtin_op_importers.cpp:533: Using kernel: (1, 1), strides: (1, 1), prepadding: (0, 0), postpadding: (0, 0), dilations: (1, 1), numOutputs: 16
[11/17/2022-03:14:15] [V] [TRT] builtin_op_importers.cpp:534: Convolution output dimensions: (1, 16, 240, 320)
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:116: Registering tensor: model_2/model/expanded_conv_project_BN/FusedBatchNormV3:0 for ONNX tensor: model_2/model/expanded_conv_project_BN/FusedBatchNormV3:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:179: model_2/model/expanded_conv_project/Conv2D [Conv] outputs: [model_2/model/expanded_conv_project_BN/FusedBatchNormV3:0 -> (1, 16, 240, 320)], 
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:103: Parsing node: model_2/model/block_1_expand/Conv2D [Conv]
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/expanded_conv_project_BN/FusedBatchNormV3:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: 317
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: 319
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:125: model_2/model/block_1_expand/Conv2D [Conv] inputs: [model_2/model/expanded_conv_project_BN/FusedBatchNormV3:0 -> (1, 16, 240, 320)], [317 -> (96, 16, 1, 1)], [319 -> (96)], 
[11/17/2022-03:14:15] [V] [TRT] builtin_op_importers.cpp:450: Convolution input dimensions: (1, 16, 240, 320)
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:141: Registering layer: model_2/model/block_1_expand/Conv2D for ONNX node: model_2/model/block_1_expand/Conv2D
[11/17/2022-03:14:15] [V] [TRT] builtin_op_importers.cpp:533: Using kernel: (1, 1), strides: (1, 1), prepadding: (0, 0), postpadding: (0, 0), dilations: (1, 1), numOutputs: 96
[11/17/2022-03:14:15] [V] [TRT] builtin_op_importers.cpp:534: Convolution output dimensions: (1, 96, 240, 320)
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:116: Registering tensor: model_2/model/block_1_expand/Conv2D:0 for ONNX tensor: model_2/model/block_1_expand/Conv2D:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:179: model_2/model/block_1_expand/Conv2D [Conv] outputs: [model_2/model/block_1_expand/Conv2D:0 -> (1, 96, 240, 320)], 
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:103: Parsing node: model_2/model/block_1_expand_relu/Relu6 [Clip]
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_1_expand/Conv2D:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_9_depthwise_relu/Relu6_min__196
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_4_expand_relu/Relu6_max__92
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:125: model_2/model/block_1_expand_relu/Relu6 [Clip] inputs: [model_2/model/block_1_expand/Conv2D:0 -> (1, 96, 240, 320)], [model_2/model/block_9_depthwise_relu/Relu6_min__196 -> ()], [model_2/model/block_4_expand_relu/Relu6_max__92 -> ()], 
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:141: Registering layer: model_2/model/block_1_expand_relu/Relu6 for ONNX node: model_2/model/block_1_expand_relu/Relu6
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:116: Registering tensor: model_2/model/block_1_expand_relu/Relu6:0 for ONNX tensor: model_2/model/block_1_expand_relu/Relu6:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:179: model_2/model/block_1_expand_relu/Relu6 [Clip] outputs: [model_2/model/block_1_expand_relu/Relu6:0 -> (1, 96, 240, 320)], 
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:103: Parsing node: model_2/model/block_1_depthwise/depthwise [Conv]
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_1_expand_relu/Relu6:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_1_depthwise/depthwise_weights_fused_bn
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_1_depthwise/depthwise_bias_fused_bn
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:125: model_2/model/block_1_depthwise/depthwise [Conv] inputs: [model_2/model/block_1_expand_relu/Relu6:0 -> (1, 96, 240, 320)], [model_2/model/block_1_depthwise/depthwise_weights_fused_bn -> (96, 1, 3, 3)], [model_2/model/block_1_depthwise/depthwise_bias_fused_bn -> (96)], 
[11/17/2022-03:14:15] [V] [TRT] builtin_op_importers.cpp:450: Convolution input dimensions: (1, 96, 240, 320)
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:141: Registering layer: model_2/model/block_1_depthwise/depthwise for ONNX node: model_2/model/block_1_depthwise/depthwise
[11/17/2022-03:14:15] [V] [TRT] builtin_op_importers.cpp:533: Using kernel: (3, 3), strides: (2, 2), prepadding: (0, 0), postpadding: (1, 1), dilations: (1, 1), numOutputs: 96
[11/17/2022-03:14:15] [V] [TRT] builtin_op_importers.cpp:534: Convolution output dimensions: (1, 96, 120, 160)
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:116: Registering tensor: model_2/model/block_1_depthwise_BN/FusedBatchNormV3:0 for ONNX tensor: model_2/model/block_1_depthwise_BN/FusedBatchNormV3:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:179: model_2/model/block_1_depthwise/depthwise [Conv] outputs: [model_2/model/block_1_depthwise_BN/FusedBatchNormV3:0 -> (1, 96, 120, 160)], 
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:103: Parsing node: model_2/model/block_1_depthwise_relu/Relu6 [Clip]
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_1_depthwise_BN/FusedBatchNormV3:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_9_depthwise_relu/Relu6_min__196
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_4_expand_relu/Relu6_max__92
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:125: model_2/model/block_1_depthwise_relu/Relu6 [Clip] inputs: [model_2/model/block_1_depthwise_BN/FusedBatchNormV3:0 -> (1, 96, 120, 160)], [model_2/model/block_9_depthwise_relu/Relu6_min__196 -> ()], [model_2/model/block_4_expand_relu/Relu6_max__92 -> ()], 
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:141: Registering layer: model_2/model/block_1_depthwise_relu/Relu6 for ONNX node: model_2/model/block_1_depthwise_relu/Relu6
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:116: Registering tensor: model_2/model/block_1_depthwise_relu/Relu6:0 for ONNX tensor: model_2/model/block_1_depthwise_relu/Relu6:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:179: model_2/model/block_1_depthwise_relu/Relu6 [Clip] outputs: [model_2/model/block_1_depthwise_relu/Relu6:0 -> (1, 96, 120, 160)], 
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:103: Parsing node: model_2/model/block_1_project/Conv2D [Conv]
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_1_depthwise_relu/Relu6:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_1_project/Conv2D_weights_fused_bn
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_1_project/Conv2D_bias_fused_bn
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:125: model_2/model/block_1_project/Conv2D [Conv] inputs: [model_2/model/block_1_depthwise_relu/Relu6:0 -> (1, 96, 120, 160)], [model_2/model/block_1_project/Conv2D_weights_fused_bn -> (24, 96, 1, 1)], [model_2/model/block_1_project/Conv2D_bias_fused_bn -> (24)], 
[11/17/2022-03:14:15] [V] [TRT] builtin_op_importers.cpp:450: Convolution input dimensions: (1, 96, 120, 160)
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:141: Registering layer: model_2/model/block_1_project/Conv2D for ONNX node: model_2/model/block_1_project/Conv2D
[11/17/2022-03:14:15] [V] [TRT] builtin_op_importers.cpp:533: Using kernel: (1, 1), strides: (1, 1), prepadding: (0, 0), postpadding: (0, 0), dilations: (1, 1), numOutputs: 24
[11/17/2022-03:14:15] [V] [TRT] builtin_op_importers.cpp:534: Convolution output dimensions: (1, 24, 120, 160)
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:116: Registering tensor: model_2/model/block_1_project_BN/FusedBatchNormV3:0 for ONNX tensor: model_2/model/block_1_project_BN/FusedBatchNormV3:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:179: model_2/model/block_1_project/Conv2D [Conv] outputs: [model_2/model/block_1_project_BN/FusedBatchNormV3:0 -> (1, 24, 120, 160)], 
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:103: Parsing node: model_2/model/block_2_expand/Conv2D [Conv]
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_1_project_BN/FusedBatchNormV3:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: 321
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: 323
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:125: model_2/model/block_2_expand/Conv2D [Conv] inputs: [model_2/model/block_1_project_BN/FusedBatchNormV3:0 -> (1, 24, 120, 160)], [321 -> (144, 24, 1, 1)], [323 -> (144)], 
[11/17/2022-03:14:15] [V] [TRT] builtin_op_importers.cpp:450: Convolution input dimensions: (1, 24, 120, 160)
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:141: Registering layer: model_2/model/block_2_expand/Conv2D for ONNX node: model_2/model/block_2_expand/Conv2D
[11/17/2022-03:14:15] [V] [TRT] builtin_op_importers.cpp:533: Using kernel: (1, 1), strides: (1, 1), prepadding: (0, 0), postpadding: (0, 0), dilations: (1, 1), numOutputs: 144
[11/17/2022-03:14:15] [V] [TRT] builtin_op_importers.cpp:534: Convolution output dimensions: (1, 144, 120, 160)
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:116: Registering tensor: model_2/model/block_2_expand/Conv2D:0 for ONNX tensor: model_2/model/block_2_expand/Conv2D:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:179: model_2/model/block_2_expand/Conv2D [Conv] outputs: [model_2/model/block_2_expand/Conv2D:0 -> (1, 144, 120, 160)], 
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:103: Parsing node: model_2/model/block_2_expand_relu/Relu6 [Clip]
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_2_expand/Conv2D:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_9_depthwise_relu/Relu6_min__196
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_4_expand_relu/Relu6_max__92
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:125: model_2/model/block_2_expand_relu/Relu6 [Clip] inputs: [model_2/model/block_2_expand/Conv2D:0 -> (1, 144, 120, 160)], [model_2/model/block_9_depthwise_relu/Relu6_min__196 -> ()], [model_2/model/block_4_expand_relu/Relu6_max__92 -> ()], 
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:141: Registering layer: model_2/model/block_2_expand_relu/Relu6 for ONNX node: model_2/model/block_2_expand_relu/Relu6
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:116: Registering tensor: model_2/model/block_2_expand_relu/Relu6:0 for ONNX tensor: model_2/model/block_2_expand_relu/Relu6:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:179: model_2/model/block_2_expand_relu/Relu6 [Clip] outputs: [model_2/model/block_2_expand_relu/Relu6:0 -> (1, 144, 120, 160)], 
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:103: Parsing node: model_2/model/block_2_depthwise/depthwise [Conv]
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_2_expand_relu/Relu6:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_2_depthwise/depthwise_weights_fused_bn
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_2_depthwise/depthwise_bias_fused_bn
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:125: model_2/model/block_2_depthwise/depthwise [Conv] inputs: [model_2/model/block_2_expand_relu/Relu6:0 -> (1, 144, 120, 160)], [model_2/model/block_2_depthwise/depthwise_weights_fused_bn -> (144, 1, 3, 3)], [model_2/model/block_2_depthwise/depthwise_bias_fused_bn -> (144)], 
[11/17/2022-03:14:15] [V] [TRT] builtin_op_importers.cpp:450: Convolution input dimensions: (1, 144, 120, 160)
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:141: Registering layer: model_2/model/block_2_depthwise/depthwise for ONNX node: model_2/model/block_2_depthwise/depthwise
[11/17/2022-03:14:15] [V] [TRT] builtin_op_importers.cpp:533: Using kernel: (3, 3), strides: (1, 1), prepadding: (1, 1), postpadding: (1, 1), dilations: (1, 1), numOutputs: 144
[11/17/2022-03:14:15] [V] [TRT] builtin_op_importers.cpp:534: Convolution output dimensions: (1, 144, 120, 160)
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:116: Registering tensor: model_2/model/block_2_depthwise_BN/FusedBatchNormV3:0 for ONNX tensor: model_2/model/block_2_depthwise_BN/FusedBatchNormV3:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:179: model_2/model/block_2_depthwise/depthwise [Conv] outputs: [model_2/model/block_2_depthwise_BN/FusedBatchNormV3:0 -> (1, 144, 120, 160)], 
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:103: Parsing node: model_2/model/block_2_depthwise_relu/Relu6 [Clip]
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_2_depthwise_BN/FusedBatchNormV3:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_9_depthwise_relu/Relu6_min__196
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_4_expand_relu/Relu6_max__92
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:125: model_2/model/block_2_depthwise_relu/Relu6 [Clip] inputs: [model_2/model/block_2_depthwise_BN/FusedBatchNormV3:0 -> (1, 144, 120, 160)], [model_2/model/block_9_depthwise_relu/Relu6_min__196 -> ()], [model_2/model/block_4_expand_relu/Relu6_max__92 -> ()], 
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:141: Registering layer: model_2/model/block_2_depthwise_relu/Relu6 for ONNX node: model_2/model/block_2_depthwise_relu/Relu6
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:116: Registering tensor: model_2/model/block_2_depthwise_relu/Relu6:0 for ONNX tensor: model_2/model/block_2_depthwise_relu/Relu6:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:179: model_2/model/block_2_depthwise_relu/Relu6 [Clip] outputs: [model_2/model/block_2_depthwise_relu/Relu6:0 -> (1, 144, 120, 160)], 
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:103: Parsing node: model_2/model/block_2_project/Conv2D [Conv]
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_2_depthwise_relu/Relu6:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_2_project/Conv2D_weights_fused_bn
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_2_project/Conv2D_bias_fused_bn
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:125: model_2/model/block_2_project/Conv2D [Conv] inputs: [model_2/model/block_2_depthwise_relu/Relu6:0 -> (1, 144, 120, 160)], [model_2/model/block_2_project/Conv2D_weights_fused_bn -> (24, 144, 1, 1)], [model_2/model/block_2_project/Conv2D_bias_fused_bn -> (24)], 
[11/17/2022-03:14:15] [V] [TRT] builtin_op_importers.cpp:450: Convolution input dimensions: (1, 144, 120, 160)
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:141: Registering layer: model_2/model/block_2_project/Conv2D for ONNX node: model_2/model/block_2_project/Conv2D
[11/17/2022-03:14:15] [V] [TRT] builtin_op_importers.cpp:533: Using kernel: (1, 1), strides: (1, 1), prepadding: (0, 0), postpadding: (0, 0), dilations: (1, 1), numOutputs: 24
[11/17/2022-03:14:15] [V] [TRT] builtin_op_importers.cpp:534: Convolution output dimensions: (1, 24, 120, 160)
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:116: Registering tensor: model_2/model/block_2_project_BN/FusedBatchNormV3:0 for ONNX tensor: model_2/model/block_2_project_BN/FusedBatchNormV3:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:179: model_2/model/block_2_project/Conv2D [Conv] outputs: [model_2/model/block_2_project_BN/FusedBatchNormV3:0 -> (1, 24, 120, 160)], 
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:103: Parsing node: model_2/model/block_2_add/add [Add]
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_1_project_BN/FusedBatchNormV3:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:119: Searching for input: model_2/model/block_2_project_BN/FusedBatchNormV3:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:125: model_2/model/block_2_add/add [Add] inputs: [model_2/model/block_1_project_BN/FusedBatchNormV3:0 -> (1, 24, 120, 160)], [model_2/model/block_2_project_BN/FusedBatchNormV3:0 -> (1, 24, 120, 160)], 
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:141: Registering layer: model_2/model/block_2_add/add for ONNX node: model_2/model/block_2_add/add
[11/17/2022-03:14:15] [V] [TRT] ImporterContext.hpp:116: Registering tensor: model_2/model/block_2_add/add:0_1 for ONNX tensor: model_2/model/block_2_add/add:0
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:179: model_2/model/block_2_add/add [Add] outputs: [model_2/model/block_2_add/add:0 -> (1, 24, 120, 160)], 
[11/17/2022-03:14:15] [V] [TRT] ModelImporter.cpp:507: Marking model_2/model/block_2_add/add:0_1 as output: model_2/model/block_2_add/add:0
 ----- Parsing of ONNX model reduced.onnx is Done ---- 
[11/17/2022-03:14:15] [V] [TRT] Applying generic optimizations to the graph for inference.
[11/17/2022-03:14:15] [V] [TRT] Original: 14 layers
[11/17/2022-03:14:15] [V] [TRT] After dead-layer removal: 14 layers
[11/17/2022-03:14:15] [V] [TRT] After Myelin optimization: 14 layers
[11/17/2022-03:14:15] [V] [TRT] After scale fusion: 14 layers
[11/17/2022-03:14:15] [V] [TRT] Fusing model_2/model/expanded_conv_depthwise/depthwise with model_2/model/expanded_conv_depthwise_relu/Relu6
[11/17/2022-03:14:15] [V] [TRT] Fusing model_2/model/block_1_expand/Conv2D with model_2/model/block_1_expand_relu/Relu6
[11/17/2022-03:14:15] [V] [TRT] Fusing model_2/model/block_1_depthwise/depthwise with model_2/model/block_1_depthwise_relu/Relu6
[11/17/2022-03:14:15] [V] [TRT] Fusing model_2/model/block_2_expand/Conv2D with model_2/model/block_2_expand_relu/Relu6
[11/17/2022-03:14:15] [V] [TRT] Fusing model_2/model/block_2_depthwise/depthwise with model_2/model/block_2_depthwise_relu/Relu6
[11/17/2022-03:14:15] [V] [TRT] Fusing model_2/model/block_2_project/Conv2D with model_2/model/block_2_add/add
[11/17/2022-03:14:15] [V] [TRT] After vertical fusions: 8 layers
[11/17/2022-03:14:15] [V] [TRT] After final dead-layer removal: 8 layers
[11/17/2022-03:14:15] [V] [TRT] After tensor merging: 8 layers
[11/17/2022-03:14:15] [V] [TRT] After concat removal: 8 layers
[11/17/2022-03:14:15] [V] [TRT] Graph construction and optimization completed in 0.00367677 seconds.
[11/17/2022-03:14:19] [V] [TRT] Constructing optimization profile number 0 [1/1].
[11/17/2022-03:14:19] [V] [TRT] *************** Autotuning format combination: Float(1,320,76800,2457600) -> Float(1,320,76800,2457600) ***************
[11/17/2022-03:14:19] [V] [TRT] model_2/model/expanded_conv_depthwise/depthwise + model_2/model/expanded_conv_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_medium_nn_v1
[11/17/2022-03:14:19] [V] [TRT] model_2/model/expanded_conv_depthwise/depthwise + model_2/model/expanded_conv_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_large_nn_v1
[11/17/2022-03:14:19] [V] [TRT] model_2/model/expanded_conv_depthwise/depthwise + model_2/model/expanded_conv_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_medium_nn_v1
[11/17/2022-03:14:19] [V] [TRT] model_2/model/expanded_conv_depthwise/depthwise + model_2/model/expanded_conv_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_small_nn_v1
[11/17/2022-03:14:19] [V] [TRT] model_2/model/expanded_conv_depthwise/depthwise + model_2/model/expanded_conv_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_medium_nn_v1
[11/17/2022-03:14:19] [V] [TRT] model_2/model/expanded_conv_depthwise/depthwise + model_2/model/expanded_conv_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_large_nn_v1
[11/17/2022-03:14:19] [V] [TRT] model_2/model/expanded_conv_depthwise/depthwise + model_2/model/expanded_conv_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_large_nn_v1
[11/17/2022-03:14:19] [V] [TRT] model_2/model/expanded_conv_depthwise/depthwise + model_2/model/expanded_conv_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_small_nn_v1
[11/17/2022-03:14:19] [V] [TRT] model_2/model/expanded_conv_depthwise/depthwise + model_2/model/expanded_conv_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_small_nn_v1
[11/17/2022-03:14:20] [V] [TRT] --------------- Timing Runner: model_2/model/expanded_conv_depthwise/depthwise + model_2/model/expanded_conv_depthwise_relu/Relu6 (FusedConvActConvolution)
[11/17/2022-03:14:20] [V] [TRT] FusedConvActConvolution has no valid tactics for this config, skipping
[11/17/2022-03:14:20] [V] [TRT] --------------- Timing Runner: model_2/model/expanded_conv_depthwise/depthwise + model_2/model/expanded_conv_depthwise_relu/Relu6 (CaskConvolution)
[11/17/2022-03:14:20] [V] [TRT] model_2/model/expanded_conv_depthwise/depthwise + model_2/model/expanded_conv_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_medium_nn_v1
[11/17/2022-03:14:20] [V] [TRT] Tactic: 1062367460111450758 time 20.8044
[11/17/2022-03:14:20] [V] [TRT] model_2/model/expanded_conv_depthwise/depthwise + model_2/model/expanded_conv_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_large_nn_v1
[11/17/2022-03:14:21] [V] [TRT] Tactic: 4337000649858996379 time 32.3899
[11/17/2022-03:14:21] [V] [TRT] model_2/model/expanded_conv_depthwise/depthwise + model_2/model/expanded_conv_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_medium_nn_v1
[11/17/2022-03:14:22] [V] [TRT] Tactic: 4501471010995462441 time 62.7859
[11/17/2022-03:14:22] [V] [TRT] model_2/model/expanded_conv_depthwise/depthwise + model_2/model/expanded_conv_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_small_nn_v1
[11/17/2022-03:14:22] [V] [TRT] Tactic: 5137655947464784826 time 31.5786
[11/17/2022-03:14:22] [V] [TRT] model_2/model/expanded_conv_depthwise/depthwise + model_2/model/expanded_conv_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_medium_nn_v1
[11/17/2022-03:14:23] [V] [TRT] Tactic: 6645123197870846056 time 31.7829
[11/17/2022-03:14:23] [V] [TRT] model_2/model/expanded_conv_depthwise/depthwise + model_2/model/expanded_conv_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_large_nn_v1
[11/17/2022-03:14:24] [V] [TRT] Tactic: -9137461792520977713 time 63.5333
[11/17/2022-03:14:24] [V] [TRT] model_2/model/expanded_conv_depthwise/depthwise + model_2/model/expanded_conv_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_large_nn_v1
[11/17/2022-03:14:24] [V] [TRT] Tactic: -6092040395344634144 time 21.0086
[11/17/2022-03:14:24] [V] [TRT] model_2/model/expanded_conv_depthwise/depthwise + model_2/model/expanded_conv_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_small_nn_v1
[11/17/2022-03:14:24] [V] [TRT] Tactic: -3456450830548107839 time 19.7843
[11/17/2022-03:14:24] [V] [TRT] model_2/model/expanded_conv_depthwise/depthwise + model_2/model/expanded_conv_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_small_nn_v1
[11/17/2022-03:14:25] [V] [TRT] Tactic: -410470605513481746 time 62.1241
[11/17/2022-03:14:26] [V] [TRT] Fastest Tactic: -3456450830548107839 Time: 19.7843
[11/17/2022-03:14:26] [V] [TRT] --------------- Timing Runner: model_2/model/expanded_conv_depthwise/depthwise + model_2/model/expanded_conv_depthwise_relu/Relu6 (CudaConvolution)
[11/17/2022-03:14:26] [V] [TRT] Tactic: 0 time 4.24082
[11/17/2022-03:14:26] [V] [TRT] Tactic: 2 time 4.23696
[11/17/2022-03:14:34] [V] [TRT] Tactic: 5 time 463.678
[11/17/2022-03:14:35] [V] [TRT] Tactic: 6 time 40.8537
[11/17/2022-03:14:35] [V] [TRT] Tactic: 57 time 4.24319
[11/17/2022-03:14:35] [V] [TRT] Fastest Tactic: 2 Time: 4.23696
[11/17/2022-03:14:35] [V] [TRT] --------------- Timing Runner: model_2/model/expanded_conv_depthwise/depthwise + model_2/model/expanded_conv_depthwise_relu/Relu6 (CudaDepthwiseConvolution)
[11/17/2022-03:14:35] [V] [TRT] Tactic: -1 time 1.56198
[11/17/2022-03:14:35] [V] [TRT] Fastest Tactic: -1 Time: 1.56198
[11/17/2022-03:14:35] [V] [TRT] >>>>>>>>>>>>>>> Chose Runner Type: CudaDepthwiseConvolution Tactic: -1
[11/17/2022-03:14:35] [V] [TRT] 
[11/17/2022-03:14:35] [V] [TRT] *************** Autotuning format combination: Float(1,320,76800,2457600) -> Float(1,320,76800,1228800) ***************
[11/17/2022-03:14:35] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_medium_nn_v1
[11/17/2022-03:14:35] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_medium_nn_v1
[11/17/2022-03:14:35] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_small_nn_v1
[11/17/2022-03:14:35] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_interior_nn_v1
[11/17/2022-03:14:35] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_medium_nn_v1
[11/17/2022-03:14:35] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_interior_nn_v1
[11/17/2022-03:14:35] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_small_nn_v1
[11/17/2022-03:14:35] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_small_nn_v1
[11/17/2022-03:14:35] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_interior_nn_v1
[11/17/2022-03:14:35] [V] [TRT] --------------- Timing Runner: model_2/model/expanded_conv_project/Conv2D (FusedConvActConvolution)
[11/17/2022-03:14:35] [V] [TRT] Tactic: 589823 time 2.47032
[11/17/2022-03:14:35] [V] [TRT] Tactic: 655359 time 3.80554
[11/17/2022-03:14:35] [V] [TRT] Tactic: 786431 time 2.60776
[11/17/2022-03:14:35] [V] [TRT] Tactic: 851967 time 6.01265
[11/17/2022-03:14:36] [V] [TRT] Tactic: 1179647 time 3.44171
[11/17/2022-03:14:36] [V] [TRT] Tactic: 1310719 time 4.02326
[11/17/2022-03:14:36] [V] [TRT] Tactic: 1376255 time 2.10742
[11/17/2022-03:14:36] [V] [TRT] Tactic: 1441791 time 4.22842
[11/17/2022-03:14:36] [V] [TRT] Tactic: 1507327 time 5.11062
[11/17/2022-03:14:36] [V] [TRT] Tactic: 1638399 time 2.74525
[11/17/2022-03:14:37] [V] [TRT] Tactic: 1835007 time 2.72639
[11/17/2022-03:14:37] [V] [TRT] Tactic: 1900543 time 4.32243
[11/17/2022-03:14:37] [V] [TRT] Tactic: 2097151 time 3.0726
[11/17/2022-03:14:37] [V] [TRT] Tactic: 2162687 time 2.21743
[11/17/2022-03:14:37] [V] [TRT] Tactic: 2293759 time 1.99194
[11/17/2022-03:14:37] [V] [TRT] Tactic: 2359295 time 2.83177
[11/17/2022-03:14:37] [V] [TRT] Tactic: 2686975 time 2.07525
[11/17/2022-03:14:38] [V] [TRT] Tactic: 3080191 time 4.6432
[11/17/2022-03:14:38] [V] [TRT] Tactic: 3342335 time 4.89306
[11/17/2022-03:14:38] [V] [TRT] Tactic: 3407871 time 2.50263
[11/17/2022-03:14:38] [V] [TRT] Tactic: 3538943 time 2.43613
[11/17/2022-03:14:38] [V] [TRT] Tactic: 3670015 time 3.30121
[11/17/2022-03:14:38] [V] [TRT] Tactic: 3932159 time 3.6312
[11/17/2022-03:14:38] [V] [TRT] Tactic: 3997695 time 2.69752
[11/17/2022-03:14:39] [V] [TRT] Tactic: 4063231 time 6.035
[11/17/2022-03:14:39] [V] [TRT] Tactic: 4194303 time 2.56061
[11/17/2022-03:14:39] [V] [TRT] Tactic: 4259839 time 3.03868
[11/17/2022-03:14:39] [V] [TRT] Tactic: 4325375 time 2.40322
[11/17/2022-03:14:39] [V] [TRT] Tactic: 4521983 time 2.25523
[11/17/2022-03:14:39] [V] [TRT] Tactic: 4587519 time 2.56706
[11/17/2022-03:14:40] [V] [TRT] Tactic: 4653055 time 4.04995
[11/17/2022-03:14:40] [V] [TRT] Tactic: 4915199 time 2.57604
[11/17/2022-03:14:40] [V] [TRT] Tactic: 4980735 time 2.30704
[11/17/2022-03:14:40] [V] [TRT] Tactic: 5177343 time 3.41317
[11/17/2022-03:14:40] [V] [TRT] Tactic: 5242879 time 2.75726
[11/17/2022-03:14:40] [V] [TRT] Tactic: 5373951 time 3.94835
[11/17/2022-03:14:40] [V] [TRT] Tactic: 5439487 time 4.36356
[11/17/2022-03:14:41] [V] [TRT] Tactic: 5570559 time 4.9314
[11/17/2022-03:14:41] [V] [TRT] Tactic: 5636095 time 6.05423
[11/17/2022-03:14:41] [V] [TRT] Tactic: 5701631 time 2.19619
[11/17/2022-03:14:41] [V] [TRT] Tactic: 5767167 time 7.90926
[11/17/2022-03:14:42] [V] [TRT] Tactic: 5832703 time 2.45482
[11/17/2022-03:14:42] [V] [TRT] Tactic: 5898239 time 3.01971
[11/17/2022-03:14:42] [V] [TRT] Tactic: 6029311 time 2.08109
[11/17/2022-03:14:42] [V] [TRT] Tactic: 6225919 time 2.5867
[11/17/2022-03:14:42] [V] [TRT] Tactic: 6291455 time 3.42376
[11/17/2022-03:14:42] [V] [TRT] Tactic: 6422527 time 4.08864
[11/17/2022-03:14:42] [V] [TRT] Tactic: 6750207 time 2.48579
[11/17/2022-03:14:43] [V] [TRT] Tactic: 6815743 time 4.24457
[11/17/2022-03:14:43] [V] [TRT] Tactic: 6946815 time 4.13033
[11/17/2022-03:14:43] [V] [TRT] Tactic: 7012351 time 3.07396
[11/17/2022-03:14:43] [V] [TRT] Tactic: 7077887 time 2.45343
[11/17/2022-03:14:43] [V] [TRT] Tactic: 7143423 time 2.97749
[11/17/2022-03:14:43] [V] [TRT] Tactic: 7208959 time 2.82579
[11/17/2022-03:14:43] [V] [TRT] Tactic: 7340031 time 2.96062
[11/17/2022-03:14:44] [V] [TRT] Tactic: 7405567 time 3.20136
[11/17/2022-03:14:44] [V] [TRT] Tactic: 7536639 time 2.76958
[11/17/2022-03:14:44] [V] [TRT] Tactic: 7602175 time 2.42394
[11/17/2022-03:14:44] [V] [TRT] Tactic: 7733247 time 3.07145
[11/17/2022-03:14:44] [V] [TRT] Tactic: 7798783 time 2.60543
[11/17/2022-03:14:44] [V] [TRT] Tactic: 8191999 time 3.00492
[11/17/2022-03:14:44] [V] [TRT] Tactic: 8257535 time 2.56671
[11/17/2022-03:14:45] [V] [TRT] Tactic: 8323071 time 2.77653
[11/17/2022-03:14:45] [V] [TRT] Tactic: 8650751 time 2.78812
[11/17/2022-03:14:45] [V] [TRT] Tactic: 8716287 time 3.8568
[11/17/2022-03:14:45] [V] [TRT] Tactic: 9109503 time 3.36422
[11/17/2022-03:14:45] [V] [TRT] Tactic: 9568255 time 2.56047
[11/17/2022-03:14:45] [V] [TRT] Tactic: 9895935 time 2.56262
[11/17/2022-03:14:45] [V] [TRT] Tactic: 10223615 time 2.09186
[11/17/2022-03:14:45] [V] [TRT] Tactic: 10354687 time 2.49396
[11/17/2022-03:14:46] [V] [TRT] Tactic: 10551295 time 2.31046
[11/17/2022-03:14:46] [V] [TRT] Tactic: 10747903 time 3.03736
[11/17/2022-03:14:46] [V] [TRT] Tactic: 10944511 time 2.2638
[11/17/2022-03:14:46] [V] [TRT] Fastest Tactic: 2293759 Time: 1.99194
[11/17/2022-03:14:46] [V] [TRT] --------------- Timing Runner: model_2/model/expanded_conv_project/Conv2D (CaskConvolution)
[11/17/2022-03:14:46] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_medium_nn_v1
[11/17/2022-03:14:46] [V] [TRT] Tactic: 1062367460111450758 time 0.94738
[11/17/2022-03:14:46] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_medium_nn_v1
[11/17/2022-03:14:46] [V] [TRT] Tactic: 4501471010995462441 time 2.58865
[11/17/2022-03:14:46] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_small_nn_v1
[11/17/2022-03:14:46] [V] [TRT] Tactic: 5137655947464784826 time 1.31928
[11/17/2022-03:14:46] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_interior_nn_v1
[11/17/2022-03:14:46] [V] [TRT] Tactic: 5326823351883942011 time 2.54402
[11/17/2022-03:14:46] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_medium_nn_v1
[11/17/2022-03:14:46] [V] [TRT] Tactic: 6645123197870846056 time 1.34045
[11/17/2022-03:14:46] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_interior_nn_v1
[11/17/2022-03:14:46] [V] [TRT] Tactic: -6576203419454146580 time 0.895628
[11/17/2022-03:14:46] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_small_nn_v1
[11/17/2022-03:14:46] [V] [TRT] Tactic: -3456450830548107839 time 0.912908
[11/17/2022-03:14:46] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_small_nn_v1
[11/17/2022-03:14:46] [V] [TRT] Tactic: -410470605513481746 time 2.56804
[11/17/2022-03:14:46] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_interior_nn_v1
[11/17/2022-03:14:46] [V] [TRT] Tactic: -37215280111360163 time 1.31308
[11/17/2022-03:14:46] [V] [TRT] Fastest Tactic: -6576203419454146580 Time: 0.895628
[11/17/2022-03:14:46] [V] [TRT] --------------- Timing Runner: model_2/model/expanded_conv_project/Conv2D (CudaConvolution)
[11/17/2022-03:14:46] [V] [TRT] Tactic: 0 time 1.77105
[11/17/2022-03:14:46] [V] [TRT] Tactic: 2 time 2.58074
[11/17/2022-03:14:46] [V] [TRT] Tactic: 5 time 7.45584
[11/17/2022-03:14:46] [V] [TRT] Tactic: 57 time 1.8044
[11/17/2022-03:14:46] [V] [TRT] Fastest Tactic: 0 Time: 1.77105
[11/17/2022-03:14:46] [V] [TRT] --------------- Timing Runner: model_2/model/expanded_conv_project/Conv2D (CudaDepthwiseConvolution)
[11/17/2022-03:14:46] [V] [TRT] CudaDepthwiseConvolution has no valid tactics for this config, skipping
[11/17/2022-03:14:46] [V] [TRT] >>>>>>>>>>>>>>> Chose Runner Type: CaskConvolution Tactic: -6576203419454146580
[11/17/2022-03:14:46] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_interior_nn_v1
[11/17/2022-03:14:46] [V] [TRT] 
[11/17/2022-03:14:46] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_medium_nn_v1
[11/17/2022-03:14:46] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_medium_nn_v1
[11/17/2022-03:14:46] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_small_nn_v1
[11/17/2022-03:14:46] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_interior_nn_v1
[11/17/2022-03:14:46] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_medium_nn_v1
[11/17/2022-03:14:46] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_interior_nn_v1
[11/17/2022-03:14:46] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_small_nn_v1
[11/17/2022-03:14:46] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_small_nn_v1
[11/17/2022-03:14:46] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_interior_nn_v1
[11/17/2022-03:14:46] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_interior_nn_v1
[11/17/2022-03:14:46] [V] [TRT] *************** Autotuning format combination: Float(1,320,76800,1228800) -> Float(1,320,76800,7372800) ***************
[11/17/2022-03:14:46] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_medium_nn_v1
[11/17/2022-03:14:46] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_medium_nn_v1
[11/17/2022-03:14:46] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_small_nn_v1
[11/17/2022-03:14:46] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_interior_nn_v1
[11/17/2022-03:14:46] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_medium_nn_v1
[11/17/2022-03:14:46] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_interior_nn_v1
[11/17/2022-03:14:46] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_small_nn_v1
[11/17/2022-03:14:46] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_small_nn_v1
[11/17/2022-03:14:46] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_interior_nn_v1
[11/17/2022-03:14:46] [V] [TRT] --------------- Timing Runner: model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (FusedConvActConvolution)
[11/17/2022-03:14:46] [V] [TRT] FusedConvActConvolution has no valid tactics for this config, skipping
[11/17/2022-03:14:46] [V] [TRT] --------------- Timing Runner: model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (CaskConvolution)
[11/17/2022-03:14:46] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_medium_nn_v1
[11/17/2022-03:14:47] [V] [TRT] Tactic: 1062367460111450758 time 2.24665
[11/17/2022-03:14:47] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_medium_nn_v1
[11/17/2022-03:14:47] [V] [TRT] Tactic: 4501471010995462441 time 2.31311
[11/17/2022-03:14:47] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_small_nn_v1
[11/17/2022-03:14:47] [V] [TRT] Tactic: 5137655947464784826 time 2.23112
[11/17/2022-03:14:47] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_interior_nn_v1
[11/17/2022-03:14:47] [V] [TRT] Tactic: 5326823351883942011 time 2.27131
[11/17/2022-03:14:47] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_medium_nn_v1
[11/17/2022-03:14:47] [V] [TRT] Tactic: 6645123197870846056 time 2.27379
[11/17/2022-03:14:47] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_interior_nn_v1
[11/17/2022-03:14:47] [V] [TRT] Tactic: -6576203419454146580 time 2.1729
[11/17/2022-03:14:47] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_small_nn_v1
[11/17/2022-03:14:47] [V] [TRT] Tactic: -3456450830548107839 time 2.16482
[11/17/2022-03:14:47] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_small_nn_v1
[11/17/2022-03:14:47] [V] [TRT] Tactic: -410470605513481746 time 2.29338
[11/17/2022-03:14:47] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_interior_nn_v1
[11/17/2022-03:14:47] [V] [TRT] Tactic: -37215280111360163 time 2.21672
[11/17/2022-03:14:47] [V] [TRT] Fastest Tactic: -3456450830548107839 Time: 2.16482
[11/17/2022-03:14:47] [V] [TRT] --------------- Timing Runner: model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (CudaConvolution)
[11/17/2022-03:14:47] [V] [TRT] Tactic: 0 time 10.1934
[11/17/2022-03:14:47] [V] [TRT] Tactic: 2 time 10.5684
[11/17/2022-03:14:48] [V] [TRT] Tactic: 5 time 18.5528
[11/17/2022-03:14:48] [V] [TRT] Tactic: 57 time 10.3617
[11/17/2022-03:14:48] [V] [TRT] Fastest Tactic: 0 Time: 10.1934
[11/17/2022-03:14:48] [V] [TRT] --------------- Timing Runner: model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (CudaDepthwiseConvolution)
[11/17/2022-03:14:48] [V] [TRT] CudaDepthwiseConvolution has no valid tactics for this config, skipping
[11/17/2022-03:14:48] [V] [TRT] >>>>>>>>>>>>>>> Chose Runner Type: CaskConvolution Tactic: -3456450830548107839
[11/17/2022-03:14:48] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_small_nn_v1
[11/17/2022-03:14:48] [V] [TRT] 
[11/17/2022-03:14:48] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_medium_nn_v1
[11/17/2022-03:14:48] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_medium_nn_v1
[11/17/2022-03:14:48] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_small_nn_v1
[11/17/2022-03:14:48] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_interior_nn_v1
[11/17/2022-03:14:48] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_medium_nn_v1
[11/17/2022-03:14:48] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_interior_nn_v1
[11/17/2022-03:14:48] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_small_nn_v1
[11/17/2022-03:14:48] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_small_nn_v1
[11/17/2022-03:14:48] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_interior_nn_v1
[11/17/2022-03:14:48] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_small_nn_v1
[11/17/2022-03:14:48] [V] [TRT] *************** Autotuning format combination: Float(1,320,76800,7372800) -> Float(1,160,19200,1843200) ***************
[11/17/2022-03:14:48] [V] [TRT] model_2/model/block_1_depthwise/depthwise + model_2/model/block_1_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_medium_nn_v1
[11/17/2022-03:14:48] [V] [TRT] model_2/model/block_1_depthwise/depthwise + model_2/model/block_1_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_large_nn_v1
[11/17/2022-03:14:48] [V] [TRT] model_2/model/block_1_depthwise/depthwise + model_2/model/block_1_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_medium_nn_v1
[11/17/2022-03:14:48] [V] [TRT] model_2/model/block_1_depthwise/depthwise + model_2/model/block_1_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_small_nn_v1
[11/17/2022-03:14:48] [V] [TRT] model_2/model/block_1_depthwise/depthwise + model_2/model/block_1_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_medium_nn_v1
[11/17/2022-03:14:48] [V] [TRT] model_2/model/block_1_depthwise/depthwise + model_2/model/block_1_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_large_nn_v1
[11/17/2022-03:14:48] [V] [TRT] model_2/model/block_1_depthwise/depthwise + model_2/model/block_1_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_large_nn_v1
[11/17/2022-03:14:48] [V] [TRT] model_2/model/block_1_depthwise/depthwise + model_2/model/block_1_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_small_nn_v1
[11/17/2022-03:14:48] [V] [TRT] model_2/model/block_1_depthwise/depthwise + model_2/model/block_1_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_small_nn_v1
[11/17/2022-03:14:48] [V] [TRT] --------------- Timing Runner: model_2/model/block_1_depthwise/depthwise + model_2/model/block_1_depthwise_relu/Relu6 (FusedConvActConvolution)
[11/17/2022-03:14:48] [V] [TRT] FusedConvActConvolution has no valid tactics for this config, skipping
[11/17/2022-03:14:48] [V] [TRT] --------------- Timing Runner: model_2/model/block_1_depthwise/depthwise + model_2/model/block_1_depthwise_relu/Relu6 (CaskConvolution)
[11/17/2022-03:14:48] [V] [TRT] model_2/model/block_1_depthwise/depthwise + model_2/model/block_1_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_medium_nn_v1
[11/17/2022-03:14:48] [V] [TRT] Tactic: 1062367460111450758 time 16.0038
[11/17/2022-03:14:48] [V] [TRT] model_2/model/block_1_depthwise/depthwise + model_2/model/block_1_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_large_nn_v1
[11/17/2022-03:14:48] [V] [TRT] Tactic: 4337000649858996379 time 24.9141
[11/17/2022-03:14:49] [V] [TRT] model_2/model/block_1_depthwise/depthwise + model_2/model/block_1_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_medium_nn_v1
[11/17/2022-03:14:49] [V] [TRT] Tactic: 4501471010995462441 time 48.1331
[11/17/2022-03:14:49] [V] [TRT] model_2/model/block_1_depthwise/depthwise + model_2/model/block_1_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_small_nn_v1
[11/17/2022-03:14:50] [V] [TRT] Tactic: 5137655947464784826 time 24.1472
[11/17/2022-03:14:50] [V] [TRT] model_2/model/block_1_depthwise/depthwise + model_2/model/block_1_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_medium_nn_v1
[11/17/2022-03:14:50] [V] [TRT] Tactic: 6645123197870846056 time 24.5407
[11/17/2022-03:14:50] [V] [TRT] model_2/model/block_1_depthwise/depthwise + model_2/model/block_1_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_large_nn_v1
[11/17/2022-03:14:51] [V] [TRT] Tactic: -9137461792520977713 time 48.7544
[11/17/2022-03:14:51] [V] [TRT] model_2/model/block_1_depthwise/depthwise + model_2/model/block_1_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_large_nn_v1
[11/17/2022-03:14:51] [V] [TRT] Tactic: -6092040395344634144 time 16.4302
[11/17/2022-03:14:51] [V] [TRT] model_2/model/block_1_depthwise/depthwise + model_2/model/block_1_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_small_nn_v1
[11/17/2022-03:14:52] [V] [TRT] Tactic: -3456450830548107839 time 15.56
[11/17/2022-03:14:52] [V] [TRT] model_2/model/block_1_depthwise/depthwise + model_2/model/block_1_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_small_nn_v1
[11/17/2022-03:14:52] [V] [TRT] Tactic: -410470605513481746 time 47.9991
[11/17/2022-03:14:52] [V] [TRT] Fastest Tactic: -3456450830548107839 Time: 15.56
[11/17/2022-03:14:52] [V] [TRT] --------------- Timing Runner: model_2/model/block_1_depthwise/depthwise + model_2/model/block_1_depthwise_relu/Relu6 (CudaConvolution)
[11/17/2022-03:14:53] [V] [TRT] Tactic: 0 time 17.4854
[11/17/2022-03:14:53] [V] [TRT] Tactic: 2 time 6.56464
[11/17/2022-03:15:15] [V] [TRT] Tactic: 5 time 1398.68
[11/17/2022-03:15:16] [V] [TRT] Tactic: 57 time 17.5423
[11/17/2022-03:15:16] [V] [TRT] Fastest Tactic: 2 Time: 6.56464
[11/17/2022-03:15:16] [V] [TRT] --------------- Timing Runner: model_2/model/block_1_depthwise/depthwise + model_2/model/block_1_depthwise_relu/Relu6 (CudaDepthwiseConvolution)
[11/17/2022-03:15:16] [V] [TRT] Tactic: -1 time 2.57658
[11/17/2022-03:15:16] [V] [TRT] Fastest Tactic: -1 Time: 2.57658
[11/17/2022-03:15:16] [V] [TRT] >>>>>>>>>>>>>>> Chose Runner Type: CudaDepthwiseConvolution Tactic: -1
[11/17/2022-03:15:16] [V] [TRT] 
[11/17/2022-03:15:16] [V] [TRT] *************** Autotuning format combination: Float(1,160,19200,1843200) -> Float(1,160,19200,460800) ***************
[11/17/2022-03:15:16] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_medium_nn_v1
[11/17/2022-03:15:16] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_medium_nn_v1
[11/17/2022-03:15:16] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_small_nn_v1
[11/17/2022-03:15:16] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_interior_nn_v1
[11/17/2022-03:15:16] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_medium_nn_v1
[11/17/2022-03:15:16] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_interior_nn_v1
[11/17/2022-03:15:16] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_small_nn_v1
[11/17/2022-03:15:16] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_small_nn_v1
[11/17/2022-03:15:16] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_interior_nn_v1
[11/17/2022-03:15:16] [V] [TRT] --------------- Timing Runner: model_2/model/block_1_project/Conv2D (FusedConvActConvolution)
[11/17/2022-03:15:16] [V] [TRT] Tactic: 589823 time 1.45104
[11/17/2022-03:15:16] [V] [TRT] Tactic: 655359 time 2.14658
[11/17/2022-03:15:16] [V] [TRT] Tactic: 786431 time 1.28525
[11/17/2022-03:15:16] [V] [TRT] Tactic: 851967 time 2.87024
[11/17/2022-03:15:16] [V] [TRT] Tactic: 1179647 time 1.5618
[11/17/2022-03:15:16] [V] [TRT] Tactic: 1310719 time 2.12608
[11/17/2022-03:15:16] [V] [TRT] Tactic: 1376255 time 1.15695
[11/17/2022-03:15:16] [V] [TRT] Tactic: 1441791 time 1.86808
[11/17/2022-03:15:17] [V] [TRT] Tactic: 1507327 time 2.5159
[11/17/2022-03:15:17] [V] [TRT] Tactic: 1638399 time 1.33831
[11/17/2022-03:15:17] [V] [TRT] Tactic: 1835007 time 1.39584
[11/17/2022-03:15:17] [V] [TRT] Tactic: 1900543 time 2.43546
[11/17/2022-03:15:17] [V] [TRT] Tactic: 2097151 time 1.37778
[11/17/2022-03:15:17] [V] [TRT] Tactic: 2162687 time 1.2769
[11/17/2022-03:15:17] [V] [TRT] Tactic: 2293759 time 1.10066
[11/17/2022-03:15:17] [V] [TRT] Tactic: 2359295 time 1.30184
[11/17/2022-03:15:17] [V] [TRT] Tactic: 2686975 time 1.1583
[11/17/2022-03:15:17] [V] [TRT] Tactic: 3080191 time 2.47562
[11/17/2022-03:15:17] [V] [TRT] Tactic: 3342335 time 2.52236
[11/17/2022-03:15:17] [V] [TRT] Tactic: 3407871 time 1.1791
[11/17/2022-03:15:18] [V] [TRT] Tactic: 3538943 time 1.19265
[11/17/2022-03:15:18] [V] [TRT] Tactic: 3670015 time 2.41135
[11/17/2022-03:15:18] [V] [TRT] Tactic: 3932159 time 2.21768
[11/17/2022-03:15:18] [V] [TRT] Tactic: 3997695 time 1.26802
[11/17/2022-03:15:18] [V] [TRT] Tactic: 4063231 time 2.77185
[11/17/2022-03:15:18] [V] [TRT] Tactic: 4194303 time 1.18538
[11/17/2022-03:15:18] [V] [TRT] Tactic: 4259839 time 1.32539
[11/17/2022-03:15:18] [V] [TRT] Tactic: 4325375 time 1.04999
[11/17/2022-03:15:18] [V] [TRT] Tactic: 4521983 time 1.16008
[11/17/2022-03:15:18] [V] [TRT] Tactic: 4587519 time 1.18184
[11/17/2022-03:15:18] [V] [TRT] Tactic: 4653055 time 1.705
[11/17/2022-03:15:18] [V] [TRT] Tactic: 4915199 time 1.153
[11/17/2022-03:15:19] [V] [TRT] Tactic: 4980735 time 1.1132
[11/17/2022-03:15:19] [V] [TRT] Tactic: 5177343 time 1.6655
[11/17/2022-03:15:19] [V] [TRT] Tactic: 5242879 time 1.3657
[11/17/2022-03:15:19] [V] [TRT] Tactic: 5373951 time 2.14485
[11/17/2022-03:15:19] [V] [TRT] Tactic: 5439487 time 2.47942
[11/17/2022-03:15:19] [V] [TRT] Tactic: 5570559 time 2.53874
[11/17/2022-03:15:19] [V] [TRT] Tactic: 5636095 time 2.77085
[11/17/2022-03:15:19] [V] [TRT] Tactic: 5701631 time 1.35384
[11/17/2022-03:15:19] [V] [TRT] Tactic: 5767167 time 5.49976
[11/17/2022-03:15:20] [V] [TRT] Tactic: 5832703 time 1.19575
[11/17/2022-03:15:20] [V] [TRT] Tactic: 5898239 time 1.3336
[11/17/2022-03:15:20] [V] [TRT] Tactic: 6029311 time 1.25804
[11/17/2022-03:15:20] [V] [TRT] Tactic: 6225919 time 1.32555
[11/17/2022-03:15:20] [V] [TRT] Tactic: 6291455 time 1.55791
[11/17/2022-03:15:20] [V] [TRT] Tactic: 6422527 time 2.29629
[11/17/2022-03:15:20] [V] [TRT] Tactic: 6750207 time 1.30376
[11/17/2022-03:15:20] [V] [TRT] Tactic: 6815743 time 2.35462
[11/17/2022-03:15:20] [V] [TRT] Tactic: 6946815 time 2.26456
[11/17/2022-03:15:20] [V] [TRT] Tactic: 7012351 time 1.38196
[11/17/2022-03:15:20] [V] [TRT] Tactic: 7077887 time 1.21211
[11/17/2022-03:15:20] [V] [TRT] Tactic: 7143423 time 1.81726
[11/17/2022-03:15:21] [V] [TRT] Tactic: 7208959 time 1.46904
[11/17/2022-03:15:21] [V] [TRT] Tactic: 7340031 time 1.35676
[11/17/2022-03:15:21] [V] [TRT] Tactic: 7405567 time 1.57124
[11/17/2022-03:15:21] [V] [TRT] Tactic: 7536639 time 1.89578
[11/17/2022-03:15:21] [V] [TRT] Tactic: 7602175 time 1.20121
[11/17/2022-03:15:21] [V] [TRT] Tactic: 7733247 time 1.60436
[11/17/2022-03:15:21] [V] [TRT] Tactic: 7798783 time 1.288
[11/17/2022-03:15:21] [V] [TRT] Tactic: 8191999 time 1.63106
[11/17/2022-03:15:21] [V] [TRT] Tactic: 8257535 time 1.24816
[11/17/2022-03:15:21] [V] [TRT] Tactic: 8323071 time 1.41417
[11/17/2022-03:15:21] [V] [TRT] Tactic: 8650751 time 1.47798
[11/17/2022-03:15:21] [V] [TRT] Tactic: 8716287 time 2.31827
[11/17/2022-03:15:22] [V] [TRT] Tactic: 9109503 time 1.5163
[11/17/2022-03:15:22] [V] [TRT] Tactic: 9568255 time 1.15628
[11/17/2022-03:15:22] [V] [TRT] Tactic: 9895935 time 1.21898
[11/17/2022-03:15:22] [V] [TRT] Tactic: 10223615 time 1.17325
[11/17/2022-03:15:22] [V] [TRT] Tactic: 10354687 time 1.21345
[11/17/2022-03:15:22] [V] [TRT] Tactic: 10551295 time 1.237
[11/17/2022-03:15:22] [V] [TRT] Tactic: 10747903 time 1.37746
[11/17/2022-03:15:22] [V] [TRT] Tactic: 10944511 time 1.11018
[11/17/2022-03:15:22] [V] [TRT] Fastest Tactic: 4325375 Time: 1.04999
[11/17/2022-03:15:22] [V] [TRT] --------------- Timing Runner: model_2/model/block_1_project/Conv2D (CaskConvolution)
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_medium_nn_v1
[11/17/2022-03:15:22] [V] [TRT] Tactic: 1062367460111450758 time 0.473976
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_medium_nn_v1
[11/17/2022-03:15:22] [V] [TRT] Tactic: 4501471010995462441 time 1.28352
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_small_nn_v1
[11/17/2022-03:15:22] [V] [TRT] Tactic: 5137655947464784826 time 0.66898
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_interior_nn_v1
[11/17/2022-03:15:22] [V] [TRT] Tactic: 5326823351883942011 time 1.24965
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_medium_nn_v1
[11/17/2022-03:15:22] [V] [TRT] Tactic: 6645123197870846056 time 0.682264
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_interior_nn_v1
[11/17/2022-03:15:22] [V] [TRT] Tactic: -6576203419454146580 time 0.445512
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_small_nn_v1
[11/17/2022-03:15:22] [V] [TRT] Tactic: -3456450830548107839 time 0.463524
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_small_nn_v1
[11/17/2022-03:15:22] [V] [TRT] Tactic: -410470605513481746 time 1.26933
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_interior_nn_v1
[11/17/2022-03:15:22] [V] [TRT] Tactic: -37215280111360163 time 0.665248
[11/17/2022-03:15:22] [V] [TRT] Fastest Tactic: -6576203419454146580 Time: 0.445512
[11/17/2022-03:15:22] [V] [TRT] --------------- Timing Runner: model_2/model/block_1_project/Conv2D (CudaConvolution)
[11/17/2022-03:15:22] [V] [TRT] Tactic: 0 time 0.9119
[11/17/2022-03:15:22] [V] [TRT] Tactic: 2 time 1.50029
[11/17/2022-03:15:22] [V] [TRT] Tactic: 5 time 2.56442
[11/17/2022-03:15:22] [V] [TRT] Tactic: 57 time 0.779432
[11/17/2022-03:15:22] [V] [TRT] Fastest Tactic: 57 Time: 0.779432
[11/17/2022-03:15:22] [V] [TRT] --------------- Timing Runner: model_2/model/block_1_project/Conv2D (CudaDepthwiseConvolution)
[11/17/2022-03:15:22] [V] [TRT] CudaDepthwiseConvolution has no valid tactics for this config, skipping
[11/17/2022-03:15:22] [V] [TRT] >>>>>>>>>>>>>>> Chose Runner Type: CaskConvolution Tactic: -6576203419454146580
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_interior_nn_v1
[11/17/2022-03:15:22] [V] [TRT] 
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_medium_nn_v1
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_medium_nn_v1
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_small_nn_v1
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_interior_nn_v1
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_medium_nn_v1
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_interior_nn_v1
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_small_nn_v1
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_small_nn_v1
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_interior_nn_v1
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_interior_nn_v1
[11/17/2022-03:15:22] [V] [TRT] *************** Autotuning format combination: Float(1,160,19200,460800) -> Float(1,160,19200,2764800) ***************
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_medium_nn_v1
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_medium_nn_v1
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_small_nn_v1
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_interior_nn_v1
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_medium_nn_v1
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_interior_nn_v1
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_small_nn_v1
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_small_nn_v1
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_interior_nn_v1
[11/17/2022-03:15:22] [V] [TRT] --------------- Timing Runner: model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (FusedConvActConvolution)
[11/17/2022-03:15:22] [V] [TRT] FusedConvActConvolution has no valid tactics for this config, skipping
[11/17/2022-03:15:22] [V] [TRT] --------------- Timing Runner: model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (CaskConvolution)
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_medium_nn_v1
[11/17/2022-03:15:22] [V] [TRT] Tactic: 1062367460111450758 time 1.07839
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_medium_nn_v1
[11/17/2022-03:15:22] [V] [TRT] Tactic: 4501471010995462441 time 1.24254
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_small_nn_v1
[11/17/2022-03:15:22] [V] [TRT] Tactic: 5137655947464784826 time 0.957656
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_interior_nn_v1
[11/17/2022-03:15:22] [V] [TRT] Tactic: 5326823351883942011 time 1.21079
[11/17/2022-03:15:22] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_medium_nn_v1
[11/17/2022-03:15:23] [V] [TRT] Tactic: 6645123197870846056 time 0.966116
[11/17/2022-03:15:23] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_interior_nn_v1
[11/17/2022-03:15:23] [V] [TRT] Tactic: -6576203419454146580 time 1.01696
[11/17/2022-03:15:23] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_small_nn_v1
[11/17/2022-03:15:23] [V] [TRT] Tactic: -3456450830548107839 time 1.03115
[11/17/2022-03:15:23] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_small_nn_v1
[11/17/2022-03:15:23] [V] [TRT] Tactic: -410470605513481746 time 1.2341
[11/17/2022-03:15:23] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_interior_nn_v1
[11/17/2022-03:15:23] [V] [TRT] Tactic: -37215280111360163 time 0.954068
[11/17/2022-03:15:23] [V] [TRT] Fastest Tactic: -37215280111360163 Time: 0.954068
[11/17/2022-03:15:23] [V] [TRT] --------------- Timing Runner: model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (CudaConvolution)
[11/17/2022-03:15:23] [V] [TRT] Tactic: 0 time 3.91392
[11/17/2022-03:15:23] [V] [TRT] Tactic: 2 time 4.14345
[11/17/2022-03:15:23] [V] [TRT] Tactic: 5 time 4.52671
[11/17/2022-03:15:23] [V] [TRT] Tactic: 57 time 4.09552
[11/17/2022-03:15:23] [V] [TRT] Fastest Tactic: 0 Time: 3.91392
[11/17/2022-03:15:23] [V] [TRT] --------------- Timing Runner: model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (CudaDepthwiseConvolution)
[11/17/2022-03:15:23] [V] [TRT] CudaDepthwiseConvolution has no valid tactics for this config, skipping
[11/17/2022-03:15:23] [V] [TRT] >>>>>>>>>>>>>>> Chose Runner Type: CaskConvolution Tactic: -37215280111360163
[11/17/2022-03:15:23] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_interior_nn_v1
[11/17/2022-03:15:23] [V] [TRT] 
[11/17/2022-03:15:23] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_medium_nn_v1
[11/17/2022-03:15:23] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_medium_nn_v1
[11/17/2022-03:15:23] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_small_nn_v1
[11/17/2022-03:15:23] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_interior_nn_v1
[11/17/2022-03:15:23] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_medium_nn_v1
[11/17/2022-03:15:23] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_interior_nn_v1
[11/17/2022-03:15:23] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_small_nn_v1
[11/17/2022-03:15:23] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_small_nn_v1
[11/17/2022-03:15:23] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_interior_nn_v1
[11/17/2022-03:15:23] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_interior_nn_v1
[11/17/2022-03:15:23] [V] [TRT] *************** Autotuning format combination: Float(1,160,19200,2764800) -> Float(1,160,19200,2764800) ***************
[11/17/2022-03:15:23] [V] [TRT] model_2/model/block_2_depthwise/depthwise + model_2/model/block_2_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_medium_nn_v1
[11/17/2022-03:15:23] [V] [TRT] model_2/model/block_2_depthwise/depthwise + model_2/model/block_2_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_large_nn_v1
[11/17/2022-03:15:23] [V] [TRT] model_2/model/block_2_depthwise/depthwise + model_2/model/block_2_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_medium_nn_v1
[11/17/2022-03:15:23] [V] [TRT] model_2/model/block_2_depthwise/depthwise + model_2/model/block_2_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_small_nn_v1
[11/17/2022-03:15:23] [V] [TRT] model_2/model/block_2_depthwise/depthwise + model_2/model/block_2_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_medium_nn_v1
[11/17/2022-03:15:23] [V] [TRT] model_2/model/block_2_depthwise/depthwise + model_2/model/block_2_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_large_nn_v1
[11/17/2022-03:15:23] [V] [TRT] model_2/model/block_2_depthwise/depthwise + model_2/model/block_2_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_large_nn_v1
[11/17/2022-03:15:23] [V] [TRT] model_2/model/block_2_depthwise/depthwise + model_2/model/block_2_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_small_nn_v1
[11/17/2022-03:15:23] [V] [TRT] model_2/model/block_2_depthwise/depthwise + model_2/model/block_2_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_small_nn_v1
[11/17/2022-03:15:23] [V] [TRT] --------------- Timing Runner: model_2/model/block_2_depthwise/depthwise + model_2/model/block_2_depthwise_relu/Relu6 (FusedConvActConvolution)
[11/17/2022-03:15:23] [V] [TRT] FusedConvActConvolution has no valid tactics for this config, skipping
[11/17/2022-03:15:23] [V] [TRT] --------------- Timing Runner: model_2/model/block_2_depthwise/depthwise + model_2/model/block_2_depthwise_relu/Relu6 (CaskConvolution)
[11/17/2022-03:15:23] [V] [TRT] model_2/model/block_2_depthwise/depthwise + model_2/model/block_2_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_medium_nn_v1
[11/17/2022-03:15:23] [V] [TRT] Tactic: 1062367460111450758 time 23.6251
[11/17/2022-03:15:23] [V] [TRT] model_2/model/block_2_depthwise/depthwise + model_2/model/block_2_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_large_nn_v1
[11/17/2022-03:15:24] [V] [TRT] Tactic: 4337000649858996379 time 37.0668
[11/17/2022-03:15:24] [V] [TRT] model_2/model/block_2_depthwise/depthwise + model_2/model/block_2_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_medium_nn_v1
[11/17/2022-03:15:25] [V] [TRT] Tactic: 4501471010995462441 time 71.6897
[11/17/2022-03:15:25] [V] [TRT] model_2/model/block_2_depthwise/depthwise + model_2/model/block_2_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_small_nn_v1
[11/17/2022-03:15:26] [V] [TRT] Tactic: 5137655947464784826 time 36.1269
[11/17/2022-03:15:26] [V] [TRT] model_2/model/block_2_depthwise/depthwise + model_2/model/block_2_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_medium_nn_v1
[11/17/2022-03:15:27] [V] [TRT] Tactic: 6645123197870846056 time 36.4751
[11/17/2022-03:15:27] [V] [TRT] model_2/model/block_2_depthwise/depthwise + model_2/model/block_2_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_large_nn_v1
[11/17/2022-03:15:28] [V] [TRT] Tactic: -9137461792520977713 time 72.1057
[11/17/2022-03:15:28] [V] [TRT] model_2/model/block_2_depthwise/depthwise + model_2/model/block_2_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_large_nn_v1
[11/17/2022-03:15:28] [V] [TRT] Tactic: -6092040395344634144 time 24.2312
[11/17/2022-03:15:28] [V] [TRT] model_2/model/block_2_depthwise/depthwise + model_2/model/block_2_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_small_nn_v1
[11/17/2022-03:15:29] [V] [TRT] Tactic: -3456450830548107839 time 22.6834
[11/17/2022-03:15:29] [V] [TRT] model_2/model/block_2_depthwise/depthwise + model_2/model/block_2_depthwise_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_small_nn_v1
[11/17/2022-03:15:30] [V] [TRT] Tactic: -410470605513481746 time 71.3486
[11/17/2022-03:15:30] [V] [TRT] Fastest Tactic: -3456450830548107839 Time: 22.6834
[11/17/2022-03:15:30] [V] [TRT] --------------- Timing Runner: model_2/model/block_2_depthwise/depthwise + model_2/model/block_2_depthwise_relu/Relu6 (CudaConvolution)
[11/17/2022-03:15:30] [V] [TRT] Tactic: 0 time 4.73575
[11/17/2022-03:15:30] [V] [TRT] Tactic: 2 time 4.73403
[11/17/2022-03:15:39] [V] [TRT] Tactic: 5 time 586.264
[11/17/2022-03:15:40] [V] [TRT] Tactic: 6 time 46.0392
[11/17/2022-03:15:40] [V] [TRT] Tactic: 57 time 4.76654
[11/17/2022-03:15:40] [V] [TRT] Fastest Tactic: 2 Time: 4.73403
[11/17/2022-03:15:40] [V] [TRT] --------------- Timing Runner: model_2/model/block_2_depthwise/depthwise + model_2/model/block_2_depthwise_relu/Relu6 (CudaDepthwiseConvolution)
[11/17/2022-03:15:40] [V] [TRT] Tactic: -1 time 1.64402
[11/17/2022-03:15:40] [V] [TRT] Fastest Tactic: -1 Time: 1.64402
[11/17/2022-03:15:40] [V] [TRT] >>>>>>>>>>>>>>> Chose Runner Type: CudaDepthwiseConvolution Tactic: -1
[11/17/2022-03:15:40] [V] [TRT] 
[11/17/2022-03:15:40] [V] [TRT] *************** Autotuning format combination: Float(1,160,19200,2764800), Float(1,160,19200,460800) -> Float(1,160,19200,460800) ***************
[11/17/2022-03:15:40] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_medium_nn_v1
[11/17/2022-03:15:40] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_medium_nn_v1
[11/17/2022-03:15:40] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_small_nn_v1
[11/17/2022-03:15:40] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_interior_nn_v1
[11/17/2022-03:15:40] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_medium_nn_v1
[11/17/2022-03:15:40] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_interior_nn_v1
[11/17/2022-03:15:40] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_small_nn_v1
[11/17/2022-03:15:40] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_small_nn_v1
[11/17/2022-03:15:40] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_interior_nn_v1
[11/17/2022-03:15:40] [V] [TRT] --------------- Timing Runner: model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (CaskConvolution)
[11/17/2022-03:15:40] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_medium_nn_v1
[11/17/2022-03:15:40] [V] [TRT] Tactic: 1062367460111450758 time 0.679148
[11/17/2022-03:15:40] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_medium_nn_v1
[11/17/2022-03:15:40] [V] [TRT] Tactic: 4501471010995462441 time 1.81024
[11/17/2022-03:15:40] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_small_nn_v1
[11/17/2022-03:15:40] [V] [TRT] Tactic: 5137655947464784826 time 0.9426
[11/17/2022-03:15:40] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_interior_nn_v1
[11/17/2022-03:15:40] [V] [TRT] Tactic: 5326823351883942011 time 1.76559
[11/17/2022-03:15:40] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_medium_nn_v1
[11/17/2022-03:15:41] [V] [TRT] Tactic: 6645123197870846056 time 0.96598
[11/17/2022-03:15:41] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_interior_nn_v1
[11/17/2022-03:15:41] [V] [TRT] Tactic: -6576203419454146580 time 0.647376
[11/17/2022-03:15:41] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_small_nn_v1
[11/17/2022-03:15:41] [V] [TRT] Tactic: -3456450830548107839 time 0.656708
[11/17/2022-03:15:41] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_small_nn_v1
[11/17/2022-03:15:41] [V] [TRT] Tactic: -410470605513481746 time 1.79171
[11/17/2022-03:15:41] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_interior_nn_v1
[11/17/2022-03:15:41] [V] [TRT] Tactic: -37215280111360163 time 0.942348
[11/17/2022-03:15:41] [V] [TRT] Fastest Tactic: -6576203419454146580 Time: 0.647376
[11/17/2022-03:15:41] [V] [TRT] --------------- Timing Runner: model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (CudaConvolution)
[11/17/2022-03:15:41] [V] [TRT] Tactic: 0 time 1.34503
[11/17/2022-03:15:41] [V] [TRT] Tactic: 2 time 2.15486
[11/17/2022-03:15:41] [V] [TRT] Tactic: 5 time 3.66999
[11/17/2022-03:15:41] [V] [TRT] Tactic: 57 time 1.11165
[11/17/2022-03:15:41] [V] [TRT] Fastest Tactic: 57 Time: 1.11165
[11/17/2022-03:15:41] [V] [TRT] >>>>>>>>>>>>>>> Chose Runner Type: CaskConvolution Tactic: -6576203419454146580
[11/17/2022-03:15:41] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_interior_nn_v1
[11/17/2022-03:15:41] [V] [TRT] 
[11/17/2022-03:15:41] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_medium_nn_v1
[11/17/2022-03:15:41] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_medium_nn_v1
[11/17/2022-03:15:41] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_small_nn_v1
[11/17/2022-03:15:41] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_interior_nn_v1
[11/17/2022-03:15:41] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_medium_nn_v1
[11/17/2022-03:15:41] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_interior_nn_v1
[11/17/2022-03:15:41] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_small_nn_v1
[11/17/2022-03:15:41] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x128_relu_small_nn_v1
[11/17/2022-03:15:41] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_interior_nn_v1
[11/17/2022-03:15:41] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_interior_nn_v1
[11/17/2022-03:15:41] [V] [TRT] Formats and tactics selection completed in 81.556 seconds.
[11/17/2022-03:15:41] [V] [TRT] After reformat layers: 8 layers
[11/17/2022-03:15:41] [V] [TRT] Block size 1048576000
[11/17/2022-03:15:41] [V] [TRT] Block size 29491200
[11/17/2022-03:15:41] [V] [TRT] Block size 11059200
[11/17/2022-03:15:41] [V] [TRT] Block size 1843200
[11/17/2022-03:15:41] [V] [TRT] Total Activation Memory: 1090969600
[11/17/2022-03:15:41] [I] [TRT] Detected 1 inputs and 1 output network tensors.
[11/17/2022-03:15:41] [V] [TRT] model_2/model/expanded_conv_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_interior_nn_v1
[11/17/2022-03:15:41] [V] [TRT] model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_small_nn_v1
[11/17/2022-03:15:41] [V] [TRT] model_2/model/block_1_project/Conv2D (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_interior_nn_v1
[11/17/2022-03:15:41] [V] [TRT] model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 (scudnn) Set Tactic Name: maxwell_scudnn_128x64_relu_interior_nn_v1
[11/17/2022-03:15:41] [V] [TRT] model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add (scudnn) Set Tactic Name: maxwell_scudnn_128x32_relu_interior_nn_v1
[11/17/2022-03:15:41] [V] [TRT] Layer: model_2/model/expanded_conv_depthwise/depthwise + model_2/model/expanded_conv_depthwise_relu/Relu6 Weights: 1152 HostPersistent: 0 DevicePersistent: 0
[11/17/2022-03:15:41] [V] [TRT] Layer: model_2/model/expanded_conv_project/Conv2D Weights: 0 HostPersistent: 3200 DevicePersistent: 463360
[11/17/2022-03:15:41] [V] [TRT] Layer: model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6 Weights: 0 HostPersistent: 1664 DevicePersistent: 467456
[11/17/2022-03:15:41] [V] [TRT] Layer: model_2/model/block_1_depthwise/depthwise + model_2/model/block_1_depthwise_relu/Relu6 Weights: 3456 HostPersistent: 0 DevicePersistent: 0
[11/17/2022-03:15:41] [V] [TRT] Layer: model_2/model/block_1_project/Conv2D Weights: 0 HostPersistent: 3200 DevicePersistent: 124928
[11/17/2022-03:15:41] [V] [TRT] Layer: model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6 Weights: 0 HostPersistent: 3200 DevicePersistent: 130048
[11/17/2022-03:15:41] [V] [TRT] Layer: model_2/model/block_2_depthwise/depthwise + model_2/model/block_2_depthwise_relu/Relu6 Weights: 5184 HostPersistent: 0 DevicePersistent: 0
[11/17/2022-03:15:41] [V] [TRT] Layer: model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add Weights: 0 HostPersistent: 3200 DevicePersistent: 129536
[11/17/2022-03:15:41] [V] [TRT] Total Host Persistent Memory: 14464
[11/17/2022-03:15:41] [V] [TRT] Total Device Persistent Memory: 1315328
[11/17/2022-03:15:41] [V] [TRT] Total Weight Memory: 9792
[11/17/2022-03:15:41] [V] [TRT] Builder timing cache: created 8 entries, 0 hit(s)
[11/17/2022-03:15:41] [V] [TRT] Engine generation completed in 85.4082 seconds.
[11/17/2022-03:15:41] [V] [TRT] Engine Layer Information:
[11/17/2022-03:15:41] [V] [TRT] Layer(DepthwiseConvolution): model_2/model/expanded_conv_depthwise/depthwise + model_2/model/expanded_conv_depthwise_relu/Relu6, Tactic: -1, model_2/model/Conv1_relu/Relu6:0[Float(32,240,320)] -> model_2/model/expanded_conv_depthwise_relu/Relu6:0[Float(32,240,320)]
[11/17/2022-03:15:41] [V] [TRT] Layer(scudnn): model_2/model/expanded_conv_project/Conv2D, Tactic: -6576203419454146580, model_2/model/expanded_conv_depthwise_relu/Relu6:0[Float(32,240,320)] -> model_2/model/expanded_conv_project_BN/FusedBatchNormV3:0[Float(16,240,320)]
[11/17/2022-03:15:41] [V] [TRT] Layer(scudnn): model_2/model/block_1_expand/Conv2D + model_2/model/block_1_expand_relu/Relu6, Tactic: -3456450830548107839, model_2/model/expanded_conv_project_BN/FusedBatchNormV3:0[Float(16,240,320)] -> model_2/model/block_1_expand_relu/Relu6:0[Float(96,240,320)]
[11/17/2022-03:15:41] [V] [TRT] Layer(DepthwiseConvolution): model_2/model/block_1_depthwise/depthwise + model_2/model/block_1_depthwise_relu/Relu6, Tactic: -1, model_2/model/block_1_expand_relu/Relu6:0[Float(96,240,320)] -> model_2/model/block_1_depthwise_relu/Relu6:0[Float(96,120,160)]
[11/17/2022-03:15:41] [V] [TRT] Layer(scudnn): model_2/model/block_1_project/Conv2D, Tactic: -6576203419454146580, model_2/model/block_1_depthwise_relu/Relu6:0[Float(96,120,160)] -> model_2/model/block_1_project_BN/FusedBatchNormV3:0[Float(24,120,160)]
[11/17/2022-03:15:41] [V] [TRT] Layer(scudnn): model_2/model/block_2_expand/Conv2D + model_2/model/block_2_expand_relu/Relu6, Tactic: -37215280111360163, model_2/model/block_1_project_BN/FusedBatchNormV3:0[Float(24,120,160)] -> model_2/model/block_2_expand_relu/Relu6:0[Float(144,120,160)]
[11/17/2022-03:15:41] [V] [TRT] Layer(DepthwiseConvolution): model_2/model/block_2_depthwise/depthwise + model_2/model/block_2_depthwise_relu/Relu6, Tactic: -1, model_2/model/block_2_expand_relu/Relu6:0[Float(144,120,160)] -> model_2/model/block_2_depthwise_relu/Relu6:0[Float(144,120,160)]
[11/17/2022-03:15:41] [V] [TRT] Layer(scudnn): model_2/model/block_2_project/Conv2D + model_2/model/block_2_add/add, Tactic: -6576203419454146580, model_2/model/block_2_depthwise_relu/Relu6:0[Float(144,120,160)], model_2/model/block_1_project_BN/FusedBatchNormV3:0[Float(24,120,160)] -> model_2/model/block_2_add/add:0[Float(24,120,160)]
[11/17/2022-03:15:41] [I] Starting inference threads
[11/17/2022-03:15:44] [I] Warmup completed 0 queries over 200 ms
[11/17/2022-03:15:44] [I] Timing trace has 0 queries over 3.02756 s
[11/17/2022-03:15:44] [I] Trace averages of 10 runs:
[11/17/2022-03:15:44] [I] Average on 10 runs - GPU latency: 11.489 ms - Host latency: 12.1616 ms (end to end 12.1711 ms, enqueue 0.431499 ms)
[11/17/2022-03:15:44] [I] Average on 10 runs - GPU latency: 11.1452 ms - Host latency: 11.8172 ms (end to end 11.8266 ms, enqueue 0.492633 ms)
[11/17/2022-03:15:44] [I] Average on 10 runs - GPU latency: 11.2667 ms - Host latency: 11.9261 ms (end to end 11.9349 ms, enqueue 0.420718 ms)
[11/17/2022-03:15:44] [I] Average on 10 runs - GPU latency: 11.5108 ms - Host latency: 12.1507 ms (end to end 12.2021 ms, enqueue 0.469208 ms)
[11/17/2022-03:15:44] [I] Average on 10 runs - GPU latency: 11.242 ms - Host latency: 11.8804 ms (end to end 11.8894 ms, enqueue 0.698914 ms)
[11/17/2022-03:15:44] [I] Average on 10 runs - GPU latency: 10.9652 ms - Host latency: 11.6003 ms (end to end 11.6098 ms, enqueue 0.847919 ms)
[11/17/2022-03:15:44] [I] Average on 10 runs - GPU latency: 11.2587 ms - Host latency: 11.9003 ms (end to end 11.9095 ms, enqueue 0.731262 ms)
[11/17/2022-03:15:44] [I] Average on 10 runs - GPU latency: 11.126 ms - Host latency: 11.7706 ms (end to end 11.7796 ms, enqueue 0.494055 ms)
[11/17/2022-03:15:44] [I] Average on 10 runs - GPU latency: 11.0775 ms - Host latency: 11.7255 ms (end to end 11.7344 ms, enqueue 0.426099 ms)
[11/17/2022-03:15:44] [I] Average on 10 runs - GPU latency: 11.4282 ms - Host latency: 12.0683 ms (end to end 12.2058 ms, enqueue 0.457312 ms)
[11/17/2022-03:15:44] [I] Average on 10 runs - GPU latency: 10.9593 ms - Host latency: 11.6005 ms (end to end 11.6099 ms, enqueue 0.590955 ms)
[11/17/2022-03:15:44] [I] Average on 10 runs - GPU latency: 11.1574 ms - Host latency: 11.7929 ms (end to end 11.8022 ms, enqueue 0.840027 ms)
[11/17/2022-03:15:44] [I] Average on 10 runs - GPU latency: 11.1592 ms - Host latency: 11.7986 ms (end to end 11.8076 ms, enqueue 0.565308 ms)
[11/17/2022-03:15:44] [I] Average on 10 runs - GPU latency: 11.0445 ms - Host latency: 11.6872 ms (end to end 11.9273 ms, enqueue 0.419495 ms)
[11/17/2022-03:15:44] [I] Average on 10 runs - GPU latency: 10.9862 ms - Host latency: 11.6233 ms (end to end 11.6325 ms, enqueue 0.416736 ms)
[11/17/2022-03:15:44] [I] Average on 10 runs - GPU latency: 11.217 ms - Host latency: 11.847 ms (end to end 11.8557 ms, enqueue 0.618835 ms)
[11/17/2022-03:15:44] [I] Average on 10 runs - GPU latency: 11.1019 ms - Host latency: 11.7327 ms (end to end 11.7418 ms, enqueue 0.953735 ms)
[11/17/2022-03:15:44] [I] Average on 10 runs - GPU latency: 11.0902 ms - Host latency: 11.7252 ms (end to end 11.735 ms, enqueue 0.672998 ms)
[11/17/2022-03:15:44] [I] Average on 10 runs - GPU latency: 11.3731 ms - Host latency: 12.0193 ms (end to end 12.0284 ms, enqueue 0.552222 ms)
[11/17/2022-03:15:44] [I] Average on 10 runs - GPU latency: 11.2764 ms - Host latency: 11.921 ms (end to end 11.9302 ms, enqueue 0.409839 ms)
[11/17/2022-03:15:44] [I] Average on 10 runs - GPU latency: 11.2276 ms - Host latency: 11.8765 ms (end to end 11.8858 ms, enqueue 0.405664 ms)
[11/17/2022-03:15:44] [I] Average on 10 runs - GPU latency: 11.1625 ms - Host latency: 11.8014 ms (end to end 11.8109 ms, enqueue 0.732617 ms)
[11/17/2022-03:15:44] [I] Average on 10 runs - GPU latency: 11.0143 ms - Host latency: 11.6508 ms (end to end 11.6603 ms, enqueue 0.493677 ms)
[11/17/2022-03:15:44] [I] Average on 10 runs - GPU latency: 11.5305 ms - Host latency: 12.1646 ms (end to end 12.1738 ms, enqueue 0.703052 ms)
[11/17/2022-03:15:44] [I] Average on 10 runs - GPU latency: 11.2611 ms - Host latency: 11.8979 ms (end to end 11.9071 ms, enqueue 0.417358 ms)
[11/17/2022-03:15:44] [I] Host Latency
[11/17/2022-03:15:44] [I] min: 11.5273 ms (end to end 11.5369 ms)
[11/17/2022-03:15:44] [I] max: 14.8665 ms (end to end 14.876 ms)
[11/17/2022-03:15:44] [I] mean: 11.8471 ms (end to end 11.872 ms)
[11/17/2022-03:15:44] [I] median: 11.834 ms (end to end 11.8433 ms)
[11/17/2022-03:15:44] [I] percentile: 14.3618 ms at 99% (end to end 14.3784 ms at 99%)
[11/17/2022-03:15:44] [I] throughput: 0 qps
[11/17/2022-03:15:44] [I] walltime: 3.02756 s
[11/17/2022-03:15:44] [I] Enqueue Time
[11/17/2022-03:15:44] [I] min: 0.26416 ms
[11/17/2022-03:15:44] [I] max: 1.93225 ms
[11/17/2022-03:15:44] [I] median: 0.44397 ms
[11/17/2022-03:15:44] [I] GPU Compute
[11/17/2022-03:15:44] [I] min: 10.895 ms
[11/17/2022-03:15:44] [I] max: 14.228 ms
[11/17/2022-03:15:44] [I] mean: 11.2043 ms
[11/17/2022-03:15:44] [I] median: 11.196 ms
[11/17/2022-03:15:44] [I] percentile: 13.695 ms at 99%
[11/17/2022-03:15:44] [I] total compute time: 2.85711 s
&&&& PASSED TensorRT.trtexec # /usr/src/tensorrt/bin/trtexec --onnx=reduced.onnx --shapes=input:1x3x480x640 --workspace=1000 --avgRuns=10 --verbose