Description
I face some problem when trying to run yolov5 on jetson nano.
I trained my own yolov5 model from yolov5s.pt on my PC and export the pt model to onnx. I want to use detectnet to load the onnx model. It can be optimized by tensorrt (I think) and then error in execution.
Procedure
export onnx
python models/export-modified.py --weights best.pt --img 640 --batch 1 # export
Starting ONNX export with onnx 1.7.0...
graph(%input : Float(1:1228800, 3:409600, 640:640, 640:1),
%model.0.conv.conv.weight : Float(32:108, 12:9, 3:3, 3:1),
%model.0.conv.bn.weight : Float(32:1),
%model.0.conv.bn.bias : Float(32:1),
%model.0.conv.bn.running_mean : Float(32:1),
%model.0.conv.bn.running_var : Float(32:1),
%model.1.conv.weight : Float(64:288, 32:9, 3:3, 3:1),
%model.1.bn.weight : Float(64:1),
%model.1.bn.bias : Float(64:1),
%model.1.bn.running_mean : Float(64:1),
%model.1.bn.running_var : Float(64:1),
%model.2.cv1.conv.weight : Float(32:64, 64:1, 1:1, 1:1),
%model.2.cv1.bn.weight : Float(32:1),
%model.2.cv1.bn.bias : Float(32:1),
%model.2.cv1.bn.running_mean : Float(32:1),
%model.2.cv1.bn.running_var : Float(32:1),
%model.2.cv2.weight : Float(32:64, 64:1, 1:1, 1:1),
%model.2.cv3.weight : Float(32:32, 32:1, 1:1, 1:1),
%model.2.cv4.conv.weight : Float(64:64, 64:1, 1:1, 1:1),
%model.2.cv4.bn.weight : Float(64:1),
%model.2.cv4.bn.bias : Float(64:1),
%model.2.cv4.bn.running_mean : Float(64:1),
%model.2.cv4.bn.running_var : Float(64:1),
%model.2.bn.weight : Float(64:1),
%model.2.bn.bias : Float(64:1),
%model.2.bn.running_mean : Float(64:1),
%model.2.bn.running_var : Float(64:1),
%model.2.m.0.cv1.conv.weight : Float(32:32, 32:1, 1:1, 1:1),
%model.2.m.0.cv1.bn.weight : Float(32:1),
%model.2.m.0.cv1.bn.bias : Float(32:1),
%model.2.m.0.cv1.bn.running_mean : Float(32:1),
%model.2.m.0.cv1.bn.running_var : Float(32:1),
%model.2.m.0.cv2.conv.weight : Float(32:288, 32:9, 3:3, 3:1),
%model.2.m.0.cv2.bn.weight : Float(32:1),
%model.2.m.0.cv2.bn.bias : Float(32:1),
%model.2.m.0.cv2.bn.running_mean : Float(32:1),
%model.2.m.0.cv2.bn.running_var : Float(32:1),
%model.3.conv.weight : Float(128:576, 64:9, 3:3, 3:1),
%model.3.bn.weight : Float(128:1),
%model.3.bn.bias : Float(128:1),
%model.3.bn.running_mean : Float(128:1),
%model.3.bn.running_var : Float(128:1),
%model.4.cv1.conv.weight : Float(64:128, 128:1, 1:1, 1:1),
%model.4.cv1.bn.weight : Float(64:1),
%model.4.cv1.bn.bias : Float(64:1),
%model.4.cv1.bn.running_mean : Float(64:1),
%model.4.cv1.bn.running_var : Float(64:1),
%model.4.cv2.weight : Float(64:128, 128:1, 1:1, 1:1),
%model.4.cv3.weight : Float(64:64, 64:1, 1:1, 1:1),
%model.4.cv4.conv.weight : Float(128:128, 128:1, 1:1, 1:1),
%model.4.cv4.bn.weight : Float(128:1),
%model.4.cv4.bn.bias : Float(128:1),
%model.4.cv4.bn.running_mean : Float(128:1),
%model.4.cv4.bn.running_var : Float(128:1),
%model.4.bn.weight : Float(128:1),
%model.4.bn.bias : Float(128:1),
%model.4.bn.running_mean : Float(128:1),
%model.4.bn.running_var : Float(128:1),
%model.4.m.0.cv1.conv.weight : Float(64:64, 64:1, 1:1, 1:1),
%model.4.m.0.cv1.bn.weight : Float(64:1),
%model.4.m.0.cv1.bn.bias : Float(64:1),
%model.4.m.0.cv1.bn.running_mean : Float(64:1),
%model.4.m.0.cv1.bn.running_var : Float(64:1),
%model.4.m.0.cv2.conv.weight : Float(64:576, 64:9, 3:3, 3:1),
%model.4.m.0.cv2.bn.weight : Float(64:1),
%model.4.m.0.cv2.bn.bias : Float(64:1),
%model.4.m.0.cv2.bn.running_mean : Float(64:1),
%model.4.m.0.cv2.bn.running_var : Float(64:1),
%model.4.m.1.cv1.conv.weight : Float(64:64, 64:1, 1:1, 1:1),
%model.4.m.1.cv1.bn.weight : Float(64:1),
%model.4.m.1.cv1.bn.bias : Float(64:1),
%model.4.m.1.cv1.bn.running_mean : Float(64:1),
%model.4.m.1.cv1.bn.running_var : Float(64:1),
%model.4.m.1.cv2.conv.weight : Float(64:576, 64:9, 3:3, 3:1),
%model.4.m.1.cv2.bn.weight : Float(64:1),
%model.4.m.1.cv2.bn.bias : Float(64:1),
%model.4.m.1.cv2.bn.running_mean : Float(64:1),
%model.4.m.1.cv2.bn.running_var : Float(64:1),
%model.4.m.2.cv1.conv.weight : Float(64:64, 64:1, 1:1, 1:1),
%model.4.m.2.cv1.bn.weight : Float(64:1),
%model.4.m.2.cv1.bn.bias : Float(64:1),
%model.4.m.2.cv1.bn.running_mean : Float(64:1),
%model.4.m.2.cv1.bn.running_var : Float(64:1),
%model.4.m.2.cv2.conv.weight : Float(64:576, 64:9, 3:3, 3:1),
%model.4.m.2.cv2.bn.weight : Float(64:1),
%model.4.m.2.cv2.bn.bias : Float(64:1),
%model.4.m.2.cv2.bn.running_mean : Float(64:1),
%model.4.m.2.cv2.bn.running_var : Float(64:1),
%model.5.conv.weight : Float(256:1152, 128:9, 3:3, 3:1),
%model.5.bn.weight : Float(256:1),
%model.5.bn.bias : Float(256:1),
%model.5.bn.running_mean : Float(256:1),
%model.5.bn.running_var : Float(256:1),
%model.6.cv1.conv.weight : Float(128:256, 256:1, 1:1, 1:1),
%model.6.cv1.bn.weight : Float(128:1),
%model.6.cv1.bn.bias : Float(128:1),
%model.6.cv1.bn.running_mean : Float(128:1),
%model.6.cv1.bn.running_var : Float(128:1),
%model.6.cv2.weight : Float(128:256, 256:1, 1:1, 1:1),
%model.6.cv3.weight : Float(128:128, 128:1, 1:1, 1:1),
%model.6.cv4.conv.weight : Float(256:256, 256:1, 1:1, 1:1),
%model.6.cv4.bn.weight : Float(256:1),
%model.6.cv4.bn.bias : Float(256:1),
%model.6.cv4.bn.running_mean : Float(256:1),
%model.6.cv4.bn.running_var : Float(256:1),
%model.6.bn.weight : Float(256:1),
%model.6.bn.bias : Float(256:1),
%model.6.bn.running_mean : Float(256:1),
%model.6.bn.running_var : Float(256:1),
%model.6.m.0.cv1.conv.weight : Float(128:128, 128:1, 1:1, 1:1),
%model.6.m.0.cv1.bn.weight : Float(128:1),
%model.6.m.0.cv1.bn.bias : Float(128:1),
%model.6.m.0.cv1.bn.running_mean : Float(128:1),
%model.6.m.0.cv1.bn.running_var : Float(128:1),
%model.6.m.0.cv2.conv.weight : Float(128:1152, 128:9, 3:3, 3:1),
%model.6.m.0.cv2.bn.weight : Float(128:1),
%model.6.m.0.cv2.bn.bias : Float(128:1),
%model.6.m.0.cv2.bn.running_mean : Float(128:1),
%model.6.m.0.cv2.bn.running_var : Float(128:1),
%model.6.m.1.cv1.conv.weight : Float(128:128, 128:1, 1:1, 1:1),
%model.6.m.1.cv1.bn.weight : Float(128:1),
%model.6.m.1.cv1.bn.bias : Float(128:1),
%model.6.m.1.cv1.bn.running_mean : Float(128:1),
%model.6.m.1.cv1.bn.running_var : Float(128:1),
%model.6.m.1.cv2.conv.weight : Float(128:1152, 128:9, 3:3, 3:1),
%model.6.m.1.cv2.bn.weight : Float(128:1),
%model.6.m.1.cv2.bn.bias : Float(128:1),
%model.6.m.1.cv2.bn.running_mean : Float(128:1),
%model.6.m.1.cv2.bn.running_var : Float(128:1),
%model.6.m.2.cv1.conv.weight : Float(128:128, 128:1, 1:1, 1:1),
%model.6.m.2.cv1.bn.weight : Float(128:1),
%model.6.m.2.cv1.bn.bias : Float(128:1),
%model.6.m.2.cv1.bn.running_mean : Float(128:1),
%model.6.m.2.cv1.bn.running_var : Float(128:1),
%model.6.m.2.cv2.conv.weight : Float(128:1152, 128:9, 3:3, 3:1),
%model.6.m.2.cv2.bn.weight : Float(128:1),
%model.6.m.2.cv2.bn.bias : Float(128:1),
%model.6.m.2.cv2.bn.running_mean : Float(128:1),
%model.6.m.2.cv2.bn.running_var : Float(128:1),
%model.7.conv.weight : Float(512:2304, 256:9, 3:3, 3:1),
%model.7.bn.weight : Float(512:1),
%model.7.bn.bias : Float(512:1),
%model.7.bn.running_mean : Float(512:1),
%model.7.bn.running_var : Float(512:1),
%model.8.cv1.conv.weight : Float(256:512, 512:1, 1:1, 1:1),
%model.8.cv1.bn.weight : Float(256:1),
%model.8.cv1.bn.bias : Float(256:1),
%model.8.cv1.bn.running_mean : Float(256:1),
%model.8.cv1.bn.running_var : Float(256:1),
%model.8.cv2.conv.weight : Float(512:1024, 1024:1, 1:1, 1:1),
%model.8.cv2.bn.weight : Float(512:1),
%model.8.cv2.bn.bias : Float(512:1),
%model.8.cv2.bn.running_mean : Float(512:1),
%model.8.cv2.bn.running_var : Float(512:1),
%model.9.cv1.conv.weight : Float(256:512, 512:1, 1:1, 1:1),
%model.9.cv1.bn.weight : Float(256:1),
%model.9.cv1.bn.bias : Float(256:1),
%model.9.cv1.bn.running_mean : Float(256:1),
%model.9.cv1.bn.running_var : Float(256:1),
%model.9.cv2.weight : Float(256:512, 512:1, 1:1, 1:1),
%model.9.cv3.weight : Float(256:256, 256:1, 1:1, 1:1),
%model.9.cv4.conv.weight : Float(512:512, 512:1, 1:1, 1:1),
%model.9.cv4.bn.weight : Float(512:1),
%model.9.cv4.bn.bias : Float(512:1),
%model.9.cv4.bn.running_mean : Float(512:1),
%model.9.cv4.bn.running_var : Float(512:1),
%model.9.bn.weight : Float(512:1),
%model.9.bn.bias : Float(512:1),
%model.9.bn.running_mean : Float(512:1),
%model.9.bn.running_var : Float(512:1),
%model.9.m.0.cv1.conv.weight : Float(256:256, 256:1, 1:1, 1:1),
%model.9.m.0.cv1.bn.weight : Float(256:1),
%model.9.m.0.cv1.bn.bias : Float(256:1),
%model.9.m.0.cv1.bn.running_mean : Float(256:1),
%model.9.m.0.cv1.bn.running_var : Float(256:1),
%model.9.m.0.cv2.conv.weight : Float(256:2304, 256:9, 3:3, 3:1),
%model.9.m.0.cv2.bn.weight : Float(256:1),
%model.9.m.0.cv2.bn.bias : Float(256:1),
%model.9.m.0.cv2.bn.running_mean : Float(256:1),
%model.9.m.0.cv2.bn.running_var : Float(256:1),
%model.10.conv.weight : Float(256:512, 512:1, 1:1, 1:1),
%model.10.bn.weight : Float(256:1),
%model.10.bn.bias : Float(256:1),
%model.10.bn.running_mean : Float(256:1),
%model.10.bn.running_var : Float(256:1),
%model.13.cv1.conv.weight : Float(128:512, 512:1, 1:1, 1:1),
%model.13.cv1.bn.weight : Float(128:1),
%model.13.cv1.bn.bias : Float(128:1),
%model.13.cv1.bn.running_mean : Float(128:1),
%model.13.cv1.bn.running_var : Float(128:1),
%model.13.cv2.weight : Float(128:512, 512:1, 1:1, 1:1),
%model.13.cv3.weight : Float(128:128, 128:1, 1:1, 1:1),
%model.13.cv4.conv.weight : Float(256:256, 256:1, 1:1, 1:1),
%model.13.cv4.bn.weight : Float(256:1),
%model.13.cv4.bn.bias : Float(256:1),
%model.13.cv4.bn.running_mean : Float(256:1),
%model.13.cv4.bn.running_var : Float(256:1),
%model.13.bn.weight : Float(256:1),
%model.13.bn.bias : Float(256:1),
%model.13.bn.running_mean : Float(256:1),
%model.13.bn.running_var : Float(256:1),
%model.13.m.0.cv1.conv.weight : Float(128:128, 128:1, 1:1, 1:1),
%model.13.m.0.cv1.bn.weight : Float(128:1),
%model.13.m.0.cv1.bn.bias : Float(128:1),
%model.13.m.0.cv1.bn.running_mean : Float(128:1),
%model.13.m.0.cv1.bn.running_var : Float(128:1),
%model.13.m.0.cv2.conv.weight : Float(128:1152, 128:9, 3:3, 3:1),
%model.13.m.0.cv2.bn.weight : Float(128:1),
%model.13.m.0.cv2.bn.bias : Float(128:1),
%model.13.m.0.cv2.bn.running_mean : Float(128:1),
%model.13.m.0.cv2.bn.running_var : Float(128:1),
%model.14.conv.weight : Float(128:256, 256:1, 1:1, 1:1),
%model.14.bn.weight : Float(128:1),
%model.14.bn.bias : Float(128:1),
%model.14.bn.running_mean : Float(128:1),
%model.14.bn.running_var : Float(128:1),
%model.17.cv1.conv.weight : Float(64:256, 256:1, 1:1, 1:1),
%model.17.cv1.bn.weight : Float(64:1),
%model.17.cv1.bn.bias : Float(64:1),
%model.17.cv1.bn.running_mean : Float(64:1),
%model.17.cv1.bn.running_var : Float(64:1),
%model.17.cv2.weight : Float(64:256, 256:1, 1:1, 1:1),
%model.17.cv3.weight : Float(64:64, 64:1, 1:1, 1:1),
%model.17.cv4.conv.weight : Float(128:128, 128:1, 1:1, 1:1),
%model.17.cv4.bn.weight : Float(128:1),
%model.17.cv4.bn.bias : Float(128:1),
%model.17.cv4.bn.running_mean : Float(128:1),
%model.17.cv4.bn.running_var : Float(128:1),
%model.17.bn.weight : Float(128:1),
%model.17.bn.bias : Float(128:1),
%model.17.bn.running_mean : Float(128:1),
%model.17.bn.running_var : Float(128:1),
%model.17.m.0.cv1.conv.weight : Float(64:64, 64:1, 1:1, 1:1),
%model.17.m.0.cv1.bn.weight : Float(64:1),
%model.17.m.0.cv1.bn.bias : Float(64:1),
%model.17.m.0.cv1.bn.running_mean : Float(64:1),
%model.17.m.0.cv1.bn.running_var : Float(64:1),
%model.17.m.0.cv2.conv.weight : Float(64:576, 64:9, 3:3, 3:1),
%model.17.m.0.cv2.bn.weight : Float(64:1),
%model.17.m.0.cv2.bn.bias : Float(64:1),
%model.17.m.0.cv2.bn.running_mean : Float(64:1),
%model.17.m.0.cv2.bn.running_var : Float(64:1),
%model.18.conv.weight : Float(128:1152, 128:9, 3:3, 3:1),
%model.18.bn.weight : Float(128:1),
%model.18.bn.bias : Float(128:1),
%model.18.bn.running_mean : Float(128:1),
%model.18.bn.running_var : Float(128:1),
%model.20.cv1.conv.weight : Float(128:256, 256:1, 1:1, 1:1),
%model.20.cv1.bn.weight : Float(128:1),
%model.20.cv1.bn.bias : Float(128:1),
%model.20.cv1.bn.running_mean : Float(128:1),
%model.20.cv1.bn.running_var : Float(128:1),
%model.20.cv2.weight : Float(128:256, 256:1, 1:1, 1:1),
%model.20.cv3.weight : Float(128:128, 128:1, 1:1, 1:1),
%model.20.cv4.conv.weight : Float(256:256, 256:1, 1:1, 1:1),
%model.20.cv4.bn.weight : Float(256:1),
%model.20.cv4.bn.bias : Float(256:1),
%model.20.cv4.bn.running_mean : Float(256:1),
%model.20.cv4.bn.running_var : Float(256:1),
%model.20.bn.weight : Float(256:1),
%model.20.bn.bias : Float(256:1),
%model.20.bn.running_mean : Float(256:1),
%model.20.bn.running_var : Float(256:1),
%model.20.m.0.cv1.conv.weight : Float(128:128, 128:1, 1:1, 1:1),
%model.20.m.0.cv1.bn.weight : Float(128:1),
%model.20.m.0.cv1.bn.bias : Float(128:1),
%model.20.m.0.cv1.bn.running_mean : Float(128:1),
%model.20.m.0.cv1.bn.running_var : Float(128:1),
%model.20.m.0.cv2.conv.weight : Float(128:1152, 128:9, 3:3, 3:1),
%model.20.m.0.cv2.bn.weight : Float(128:1),
%model.20.m.0.cv2.bn.bias : Float(128:1),
%model.20.m.0.cv2.bn.running_mean : Float(128:1),
%model.20.m.0.cv2.bn.running_var : Float(128:1),
%model.21.conv.weight : Float(256:2304, 256:9, 3:3, 3:1),
%model.21.bn.weight : Float(256:1),
%model.21.bn.bias : Float(256:1),
%model.21.bn.running_mean : Float(256:1),
%model.21.bn.running_var : Float(256:1),
%model.23.cv1.conv.weight : Float(256:512, 512:1, 1:1, 1:1),
%model.23.cv1.bn.weight : Float(256:1),
%model.23.cv1.bn.bias : Float(256:1),
%model.23.cv1.bn.running_mean : Float(256:1),
%model.23.cv1.bn.running_var : Float(256:1),
%model.23.cv2.weight : Float(256:512, 512:1, 1:1, 1:1),
%model.23.cv3.weight : Float(256:256, 256:1, 1:1, 1:1),
%model.23.cv4.conv.weight : Float(512:512, 512:1, 1:1, 1:1),
%model.23.cv4.bn.weight : Float(512:1),
%model.23.cv4.bn.bias : Float(512:1),
%model.23.cv4.bn.running_mean : Float(512:1),
%model.23.cv4.bn.running_var : Float(512:1),
%model.23.bn.weight : Float(512:1),
%model.23.bn.bias : Float(512:1),
%model.23.bn.running_mean : Float(512:1),
%model.23.bn.running_var : Float(512:1),
%model.23.m.0.cv1.conv.weight : Float(256:256, 256:1, 1:1, 1:1),
%model.23.m.0.cv1.bn.weight : Float(256:1),
%model.23.m.0.cv1.bn.bias : Float(256:1),
%model.23.m.0.cv1.bn.running_mean : Float(256:1),
%model.23.m.0.cv1.bn.running_var : Float(256:1),
%model.23.m.0.cv2.conv.weight : Float(256:2304, 256:9, 3:3, 3:1),
%model.23.m.0.cv2.bn.weight : Float(256:1),
%model.23.m.0.cv2.bn.bias : Float(256:1),
%model.23.m.0.cv2.bn.running_mean : Float(256:1),
%model.23.m.0.cv2.bn.running_var : Float(256:1),
%model.24.m.0.weight : Float(21:128, 128:1, 1:1, 1:1),
%model.24.m.0.bias : Float(21:1),
%model.24.m.1.weight : Float(21:256, 256:1, 1:1, 1:1),
%model.24.m.1.bias : Float(21:1),
%model.24.m.2.weight : Float(21:512, 512:1, 1:1, 1:1),
%model.24.m.2.bias : Float(21:1),
%1061 : Float(4:1),
%1066 : Float(4:1),
%1067 : Long(1:1),
%1068 : Long(1:1),
%1069 : Long(1:1),
%1070 : Long(1:1),
%1071 : Long(1:1),
%1072 : Long(1:1)):
%371 : Tensor = onnx::Constant[value={2}]()
%372 : Tensor = onnx::Constant[value={0}]()
%373 : Tensor = onnx::Constant[value={9223372036854775807}]()
%374 : Tensor = onnx::Constant[value={2}]()
%375 : Float(1:1228800, 3:409600, 320:1280, 640:1) = onnx::Slice(%input, %372, %373, %371, %374) # /home/ares/文档/yolov5-master-new/models/common.py:89:0
%376 : Tensor = onnx::Constant[value={3}]()
%377 : Tensor = onnx::Constant[value={0}]()
%378 : Tensor = onnx::Constant[value={9223372036854775807}]()
%379 : Tensor = onnx::Constant[value={2}]()
%380 : Float(1:1228800, 3:409600, 320:1280, 320:2) = onnx::Slice(%375, %377, %378, %376, %379) # /home/ares/文档/yolov5-master-new/models/common.py:89:0
%381 : Tensor = onnx::Constant[value={2}]()
%382 : Tensor = onnx::Constant[value={1}]()
%383 : Tensor = onnx::Constant[value={9223372036854775807}]()
%384 : Tensor = onnx::Constant[value={2}]()
%385 : Float(1:1228800, 3:409600, 320:1280, 640:1) = onnx::Slice(%input, %382, %383, %381, %384) # /home/ares/文档/yolov5-master-new/models/common.py:89:0
%386 : Tensor = onnx::Constant[value={3}]()
%387 : Tensor = onnx::Constant[value={0}]()
%388 : Tensor = onnx::Constant[value={9223372036854775807}]()
%389 : Tensor = onnx::Constant[value={2}]()
%390 : Float(1:1228800, 3:409600, 320:1280, 320:2) = onnx::Slice(%385, %387, %388, %386, %389) # /home/ares/文档/yolov5-master-new/models/common.py:89:0
%391 : Tensor = onnx::Constant[value={2}]()
%392 : Tensor = onnx::Constant[value={0}]()
%393 : Tensor = onnx::Constant[value={9223372036854775807}]()
%394 : Tensor = onnx::Constant[value={2}]()
%395 : Float(1:1228800, 3:409600, 320:1280, 640:1) = onnx::Slice(%input, %392, %393, %391, %394) # /home/ares/文档/yolov5-master-new/models/common.py:89:0
%396 : Tensor = onnx::Constant[value={3}]()
%397 : Tensor = onnx::Constant[value={1}]()
%398 : Tensor = onnx::Constant[value={9223372036854775807}]()
%399 : Tensor = onnx::Constant[value={2}]()
%400 : Float(1:1228800, 3:409600, 320:1280, 320:2) = onnx::Slice(%395, %397, %398, %396, %399) # /home/ares/文档/yolov5-master-new/models/common.py:89:0
%401 : Tensor = onnx::Constant[value={2}]()
%402 : Tensor = onnx::Constant[value={1}]()
%403 : Tensor = onnx::Constant[value={9223372036854775807}]()
%404 : Tensor = onnx::Constant[value={2}]()
%405 : Float(1:1228800, 3:409600, 320:1280, 640:1) = onnx::Slice(%input, %402, %403, %401, %404) # /home/ares/文档/yolov5-master-new/models/common.py:89:0
%406 : Tensor = onnx::Constant[value={3}]()
%407 : Tensor = onnx::Constant[value={1}]()
%408 : Tensor = onnx::Constant[value={9223372036854775807}]()
%409 : Tensor = onnx::Constant[value={2}]()
%410 : Float(1:1228800, 3:409600, 320:1280, 320:2) = onnx::Slice(%405, %407, %408, %406, %409) # /home/ares/文档/yolov5-master-new/models/common.py:89:0
%411 : Float(1:1228800, 12:102400, 320:320, 320:1) = onnx::Concat[axis=1](%380, %390, %400, %410) # /home/ares/文档/yolov5-master-new/models/common.py:89:0
%412 : Float(1:3276800, 32:102400, 320:320, 320:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%411, %model.0.conv.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%413 : Float(1:3276800, 32:102400, 320:320, 320:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%412, %model.0.conv.bn.weight, %model.0.conv.bn.bias, %model.0.conv.bn.running_mean, %model.0.conv.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%414 : Float() = onnx::Constant[value={3}]()
%415 : Float(1:3276800, 32:102400, 320:320, 320:1) = onnx::Add(%413, %414)
%416 : Tensor = onnx::Constant[value={0}]()
%417 : Tensor = onnx::Constant[value={6}]()
%418 : Float(1:3276800, 32:102400, 320:320, 320:1) = onnx::Clip(%415, %416, %417) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%419 : Float(1:3276800, 32:102400, 320:320, 320:1) = onnx::Mul(%413, %418) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%420 : Float() = onnx::Constant[value={6}]()
%421 : Float(1:3276800, 32:102400, 320:320, 320:1) = onnx::Div(%419, %420)
%422 : Float(1:1638400, 64:25600, 160:160, 160:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[2, 2]](%421, %model.1.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%423 : Float(1:1638400, 64:25600, 160:160, 160:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%422, %model.1.bn.weight, %model.1.bn.bias, %model.1.bn.running_mean, %model.1.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%424 : Float() = onnx::Constant[value={3}]()
%425 : Float(1:1638400, 64:25600, 160:160, 160:1) = onnx::Add(%423, %424)
%426 : Tensor = onnx::Constant[value={0}]()
%427 : Tensor = onnx::Constant[value={6}]()
%428 : Float(1:1638400, 64:25600, 160:160, 160:1) = onnx::Clip(%425, %426, %427) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%429 : Float(1:1638400, 64:25600, 160:160, 160:1) = onnx::Mul(%423, %428) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%430 : Float() = onnx::Constant[value={6}]()
%431 : Float(1:1638400, 64:25600, 160:160, 160:1) = onnx::Div(%429, %430)
%432 : Float(1:819200, 32:25600, 160:160, 160:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%431, %model.2.cv1.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%433 : Float(1:819200, 32:25600, 160:160, 160:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%432, %model.2.cv1.bn.weight, %model.2.cv1.bn.bias, %model.2.cv1.bn.running_mean, %model.2.cv1.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%434 : Float() = onnx::Constant[value={3}]()
%435 : Float(1:819200, 32:25600, 160:160, 160:1) = onnx::Add(%433, %434)
%436 : Tensor = onnx::Constant[value={0}]()
%437 : Tensor = onnx::Constant[value={6}]()
%438 : Float(1:819200, 32:25600, 160:160, 160:1) = onnx::Clip(%435, %436, %437) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%439 : Float(1:819200, 32:25600, 160:160, 160:1) = onnx::Mul(%433, %438) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%440 : Float() = onnx::Constant[value={6}]()
%441 : Float(1:819200, 32:25600, 160:160, 160:1) = onnx::Div(%439, %440)
%442 : Float(1:819200, 32:25600, 160:160, 160:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%441, %model.2.m.0.cv1.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%443 : Float(1:819200, 32:25600, 160:160, 160:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%442, %model.2.m.0.cv1.bn.weight, %model.2.m.0.cv1.bn.bias, %model.2.m.0.cv1.bn.running_mean, %model.2.m.0.cv1.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%444 : Float() = onnx::Constant[value={3}]()
%445 : Float(1:819200, 32:25600, 160:160, 160:1) = onnx::Add(%443, %444)
%446 : Tensor = onnx::Constant[value={0}]()
%447 : Tensor = onnx::Constant[value={6}]()
%448 : Float(1:819200, 32:25600, 160:160, 160:1) = onnx::Clip(%445, %446, %447) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%449 : Float(1:819200, 32:25600, 160:160, 160:1) = onnx::Mul(%443, %448) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%450 : Float() = onnx::Constant[value={6}]()
%451 : Float(1:819200, 32:25600, 160:160, 160:1) = onnx::Div(%449, %450)
%452 : Float(1:819200, 32:25600, 160:160, 160:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%451, %model.2.m.0.cv2.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%453 : Float(1:819200, 32:25600, 160:160, 160:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%452, %model.2.m.0.cv2.bn.weight, %model.2.m.0.cv2.bn.bias, %model.2.m.0.cv2.bn.running_mean, %model.2.m.0.cv2.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%454 : Float() = onnx::Constant[value={3}]()
%455 : Float(1:819200, 32:25600, 160:160, 160:1) = onnx::Add(%453, %454)
%456 : Tensor = onnx::Constant[value={0}]()
%457 : Tensor = onnx::Constant[value={6}]()
%458 : Float(1:819200, 32:25600, 160:160, 160:1) = onnx::Clip(%455, %456, %457) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%459 : Float(1:819200, 32:25600, 160:160, 160:1) = onnx::Mul(%453, %458) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%460 : Float() = onnx::Constant[value={6}]()
%461 : Float(1:819200, 32:25600, 160:160, 160:1) = onnx::Div(%459, %460)
%462 : Float(1:819200, 32:25600, 160:160, 160:1) = onnx::Add(%441, %461) # /home/ares/文档/yolov5-master-new/models/common.py:46:0
%463 : Float(1:819200, 32:25600, 160:160, 160:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%462, %model.2.cv3.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%464 : Float(1:819200, 32:25600, 160:160, 160:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%431, %model.2.cv2.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%465 : Float(1:1638400, 64:25600, 160:160, 160:1) = onnx::Concat[axis=1](%463, %464) # /home/ares/文档/yolov5-master-new/models/common.py:65:0
%466 : Float(1:1638400, 64:25600, 160:160, 160:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%465, %model.2.bn.weight, %model.2.bn.bias, %model.2.bn.running_mean, %model.2.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%467 : Float(1:1638400, 64:25600, 160:160, 160:1) = onnx::LeakyRelu[alpha=0.10000000000000001](%466) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1293:0
%468 : Float(1:1638400, 64:25600, 160:160, 160:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%467, %model.2.cv4.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%469 : Float(1:1638400, 64:25600, 160:160, 160:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%468, %model.2.cv4.bn.weight, %model.2.cv4.bn.bias, %model.2.cv4.bn.running_mean, %model.2.cv4.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%470 : Float() = onnx::Constant[value={3}]()
%471 : Float(1:1638400, 64:25600, 160:160, 160:1) = onnx::Add(%469, %470)
%472 : Tensor = onnx::Constant[value={0}]()
%473 : Tensor = onnx::Constant[value={6}]()
%474 : Float(1:1638400, 64:25600, 160:160, 160:1) = onnx::Clip(%471, %472, %473) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%475 : Float(1:1638400, 64:25600, 160:160, 160:1) = onnx::Mul(%469, %474) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%476 : Float() = onnx::Constant[value={6}]()
%477 : Float(1:1638400, 64:25600, 160:160, 160:1) = onnx::Div(%475, %476)
%478 : Float(1:819200, 128:6400, 80:80, 80:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[2, 2]](%477, %model.3.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%479 : Float(1:819200, 128:6400, 80:80, 80:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%478, %model.3.bn.weight, %model.3.bn.bias, %model.3.bn.running_mean, %model.3.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%480 : Float() = onnx::Constant[value={3}]()
%481 : Float(1:819200, 128:6400, 80:80, 80:1) = onnx::Add(%479, %480)
%482 : Tensor = onnx::Constant[value={0}]()
%483 : Tensor = onnx::Constant[value={6}]()
%484 : Float(1:819200, 128:6400, 80:80, 80:1) = onnx::Clip(%481, %482, %483) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%485 : Float(1:819200, 128:6400, 80:80, 80:1) = onnx::Mul(%479, %484) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%486 : Float() = onnx::Constant[value={6}]()
%487 : Float(1:819200, 128:6400, 80:80, 80:1) = onnx::Div(%485, %486)
%488 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%487, %model.4.cv1.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%489 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%488, %model.4.cv1.bn.weight, %model.4.cv1.bn.bias, %model.4.cv1.bn.running_mean, %model.4.cv1.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%490 : Float() = onnx::Constant[value={3}]()
%491 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Add(%489, %490)
%492 : Tensor = onnx::Constant[value={0}]()
%493 : Tensor = onnx::Constant[value={6}]()
%494 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Clip(%491, %492, %493) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%495 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Mul(%489, %494) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%496 : Float() = onnx::Constant[value={6}]()
%497 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Div(%495, %496)
%498 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%497, %model.4.m.0.cv1.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%499 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%498, %model.4.m.0.cv1.bn.weight, %model.4.m.0.cv1.bn.bias, %model.4.m.0.cv1.bn.running_mean, %model.4.m.0.cv1.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%500 : Float() = onnx::Constant[value={3}]()
%501 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Add(%499, %500)
%502 : Tensor = onnx::Constant[value={0}]()
%503 : Tensor = onnx::Constant[value={6}]()
%504 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Clip(%501, %502, %503) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%505 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Mul(%499, %504) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%506 : Float() = onnx::Constant[value={6}]()
%507 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Div(%505, %506)
%508 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%507, %model.4.m.0.cv2.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%509 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%508, %model.4.m.0.cv2.bn.weight, %model.4.m.0.cv2.bn.bias, %model.4.m.0.cv2.bn.running_mean, %model.4.m.0.cv2.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%510 : Float() = onnx::Constant[value={3}]()
%511 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Add(%509, %510)
%512 : Tensor = onnx::Constant[value={0}]()
%513 : Tensor = onnx::Constant[value={6}]()
%514 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Clip(%511, %512, %513) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%515 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Mul(%509, %514) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%516 : Float() = onnx::Constant[value={6}]()
%517 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Div(%515, %516)
%518 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Add(%497, %517) # /home/ares/文档/yolov5-master-new/models/common.py:46:0
%519 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%518, %model.4.m.1.cv1.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%520 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%519, %model.4.m.1.cv1.bn.weight, %model.4.m.1.cv1.bn.bias, %model.4.m.1.cv1.bn.running_mean, %model.4.m.1.cv1.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%521 : Float() = onnx::Constant[value={3}]()
%522 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Add(%520, %521)
%523 : Tensor = onnx::Constant[value={0}]()
%524 : Tensor = onnx::Constant[value={6}]()
%525 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Clip(%522, %523, %524) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%526 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Mul(%520, %525) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%527 : Float() = onnx::Constant[value={6}]()
%528 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Div(%526, %527)
%529 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%528, %model.4.m.1.cv2.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%530 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%529, %model.4.m.1.cv2.bn.weight, %model.4.m.1.cv2.bn.bias, %model.4.m.1.cv2.bn.running_mean, %model.4.m.1.cv2.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%531 : Float() = onnx::Constant[value={3}]()
%532 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Add(%530, %531)
%533 : Tensor = onnx::Constant[value={0}]()
%534 : Tensor = onnx::Constant[value={6}]()
%535 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Clip(%532, %533, %534) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%536 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Mul(%530, %535) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%537 : Float() = onnx::Constant[value={6}]()
%538 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Div(%536, %537)
%539 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Add(%518, %538) # /home/ares/文档/yolov5-master-new/models/common.py:46:0
%540 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%539, %model.4.m.2.cv1.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%541 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%540, %model.4.m.2.cv1.bn.weight, %model.4.m.2.cv1.bn.bias, %model.4.m.2.cv1.bn.running_mean, %model.4.m.2.cv1.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%542 : Float() = onnx::Constant[value={3}]()
%543 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Add(%541, %542)
%544 : Tensor = onnx::Constant[value={0}]()
%545 : Tensor = onnx::Constant[value={6}]()
%546 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Clip(%543, %544, %545) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%547 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Mul(%541, %546) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%548 : Float() = onnx::Constant[value={6}]()
%549 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Div(%547, %548)
%550 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%549, %model.4.m.2.cv2.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%551 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%550, %model.4.m.2.cv2.bn.weight, %model.4.m.2.cv2.bn.bias, %model.4.m.2.cv2.bn.running_mean, %model.4.m.2.cv2.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%552 : Float() = onnx::Constant[value={3}]()
%553 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Add(%551, %552)
%554 : Tensor = onnx::Constant[value={0}]()
%555 : Tensor = onnx::Constant[value={6}]()
%556 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Clip(%553, %554, %555) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%557 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Mul(%551, %556) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%558 : Float() = onnx::Constant[value={6}]()
%559 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Div(%557, %558)
%560 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Add(%539, %559) # /home/ares/文档/yolov5-master-new/models/common.py:46:0
%561 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%560, %model.4.cv3.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%562 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%487, %model.4.cv2.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%563 : Float(1:819200, 128:6400, 80:80, 80:1) = onnx::Concat[axis=1](%561, %562) # /home/ares/文档/yolov5-master-new/models/common.py:65:0
%564 : Float(1:819200, 128:6400, 80:80, 80:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%563, %model.4.bn.weight, %model.4.bn.bias, %model.4.bn.running_mean, %model.4.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%565 : Float(1:819200, 128:6400, 80:80, 80:1) = onnx::LeakyRelu[alpha=0.10000000000000001](%564) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1293:0
%566 : Float(1:819200, 128:6400, 80:80, 80:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%565, %model.4.cv4.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%567 : Float(1:819200, 128:6400, 80:80, 80:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%566, %model.4.cv4.bn.weight, %model.4.cv4.bn.bias, %model.4.cv4.bn.running_mean, %model.4.cv4.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%568 : Float() = onnx::Constant[value={3}]()
%569 : Float(1:819200, 128:6400, 80:80, 80:1) = onnx::Add(%567, %568)
%570 : Tensor = onnx::Constant[value={0}]()
%571 : Tensor = onnx::Constant[value={6}]()
%572 : Float(1:819200, 128:6400, 80:80, 80:1) = onnx::Clip(%569, %570, %571) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%573 : Float(1:819200, 128:6400, 80:80, 80:1) = onnx::Mul(%567, %572) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%574 : Float() = onnx::Constant[value={6}]()
%575 : Float(1:819200, 128:6400, 80:80, 80:1) = onnx::Div(%573, %574)
%576 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[2, 2]](%575, %model.5.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%577 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%576, %model.5.bn.weight, %model.5.bn.bias, %model.5.bn.running_mean, %model.5.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%578 : Float() = onnx::Constant[value={3}]()
%579 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::Add(%577, %578)
%580 : Tensor = onnx::Constant[value={0}]()
%581 : Tensor = onnx::Constant[value={6}]()
%582 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::Clip(%579, %580, %581) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%583 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::Mul(%577, %582) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%584 : Float() = onnx::Constant[value={6}]()
%585 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::Div(%583, %584)
%586 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%585, %model.6.cv1.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%587 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%586, %model.6.cv1.bn.weight, %model.6.cv1.bn.bias, %model.6.cv1.bn.running_mean, %model.6.cv1.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%588 : Float() = onnx::Constant[value={3}]()
%589 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Add(%587, %588)
%590 : Tensor = onnx::Constant[value={0}]()
%591 : Tensor = onnx::Constant[value={6}]()
%592 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Clip(%589, %590, %591) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%593 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Mul(%587, %592) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%594 : Float() = onnx::Constant[value={6}]()
%595 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Div(%593, %594)
%596 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%595, %model.6.m.0.cv1.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%597 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%596, %model.6.m.0.cv1.bn.weight, %model.6.m.0.cv1.bn.bias, %model.6.m.0.cv1.bn.running_mean, %model.6.m.0.cv1.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%598 : Float() = onnx::Constant[value={3}]()
%599 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Add(%597, %598)
%600 : Tensor = onnx::Constant[value={0}]()
%601 : Tensor = onnx::Constant[value={6}]()
%602 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Clip(%599, %600, %601) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%603 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Mul(%597, %602) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%604 : Float() = onnx::Constant[value={6}]()
%605 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Div(%603, %604)
%606 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%605, %model.6.m.0.cv2.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%607 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%606, %model.6.m.0.cv2.bn.weight, %model.6.m.0.cv2.bn.bias, %model.6.m.0.cv2.bn.running_mean, %model.6.m.0.cv2.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%608 : Float() = onnx::Constant[value={3}]()
%609 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Add(%607, %608)
%610 : Tensor = onnx::Constant[value={0}]()
%611 : Tensor = onnx::Constant[value={6}]()
%612 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Clip(%609, %610, %611) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%613 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Mul(%607, %612) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%614 : Float() = onnx::Constant[value={6}]()
%615 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Div(%613, %614)
%616 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Add(%595, %615) # /home/ares/文档/yolov5-master-new/models/common.py:46:0
%617 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%616, %model.6.m.1.cv1.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%618 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%617, %model.6.m.1.cv1.bn.weight, %model.6.m.1.cv1.bn.bias, %model.6.m.1.cv1.bn.running_mean, %model.6.m.1.cv1.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%619 : Float() = onnx::Constant[value={3}]()
%620 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Add(%618, %619)
%621 : Tensor = onnx::Constant[value={0}]()
%622 : Tensor = onnx::Constant[value={6}]()
%623 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Clip(%620, %621, %622) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%624 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Mul(%618, %623) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%625 : Float() = onnx::Constant[value={6}]()
%626 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Div(%624, %625)
%627 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%626, %model.6.m.1.cv2.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%628 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%627, %model.6.m.1.cv2.bn.weight, %model.6.m.1.cv2.bn.bias, %model.6.m.1.cv2.bn.running_mean, %model.6.m.1.cv2.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%629 : Float() = onnx::Constant[value={3}]()
%630 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Add(%628, %629)
%631 : Tensor = onnx::Constant[value={0}]()
%632 : Tensor = onnx::Constant[value={6}]()
%633 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Clip(%630, %631, %632) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%634 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Mul(%628, %633) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%635 : Float() = onnx::Constant[value={6}]()
%636 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Div(%634, %635)
%637 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Add(%616, %636) # /home/ares/文档/yolov5-master-new/models/common.py:46:0
%638 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%637, %model.6.m.2.cv1.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%639 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%638, %model.6.m.2.cv1.bn.weight, %model.6.m.2.cv1.bn.bias, %model.6.m.2.cv1.bn.running_mean, %model.6.m.2.cv1.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%640 : Float() = onnx::Constant[value={3}]()
%641 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Add(%639, %640)
%642 : Tensor = onnx::Constant[value={0}]()
%643 : Tensor = onnx::Constant[value={6}]()
%644 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Clip(%641, %642, %643) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%645 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Mul(%639, %644) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%646 : Float() = onnx::Constant[value={6}]()
%647 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Div(%645, %646)
%648 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%647, %model.6.m.2.cv2.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%649 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%648, %model.6.m.2.cv2.bn.weight, %model.6.m.2.cv2.bn.bias, %model.6.m.2.cv2.bn.running_mean, %model.6.m.2.cv2.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%650 : Float() = onnx::Constant[value={3}]()
%651 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Add(%649, %650)
%652 : Tensor = onnx::Constant[value={0}]()
%653 : Tensor = onnx::Constant[value={6}]()
%654 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Clip(%651, %652, %653) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%655 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Mul(%649, %654) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%656 : Float() = onnx::Constant[value={6}]()
%657 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Div(%655, %656)
%658 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Add(%637, %657) # /home/ares/文档/yolov5-master-new/models/common.py:46:0
%659 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%658, %model.6.cv3.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%660 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%585, %model.6.cv2.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%661 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::Concat[axis=1](%659, %660) # /home/ares/文档/yolov5-master-new/models/common.py:65:0
%662 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%661, %model.6.bn.weight, %model.6.bn.bias, %model.6.bn.running_mean, %model.6.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%663 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::LeakyRelu[alpha=0.10000000000000001](%662) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1293:0
%664 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%663, %model.6.cv4.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%665 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%664, %model.6.cv4.bn.weight, %model.6.cv4.bn.bias, %model.6.cv4.bn.running_mean, %model.6.cv4.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%666 : Float() = onnx::Constant[value={3}]()
%667 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::Add(%665, %666)
%668 : Tensor = onnx::Constant[value={0}]()
%669 : Tensor = onnx::Constant[value={6}]()
%670 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::Clip(%667, %668, %669) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%671 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::Mul(%665, %670) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%672 : Float() = onnx::Constant[value={6}]()
%673 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::Div(%671, %672)
%674 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[2, 2]](%673, %model.7.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%675 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%674, %model.7.bn.weight, %model.7.bn.bias, %model.7.bn.running_mean, %model.7.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%676 : Float() = onnx::Constant[value={3}]()
%677 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::Add(%675, %676)
%678 : Tensor = onnx::Constant[value={0}]()
%679 : Tensor = onnx::Constant[value={6}]()
%680 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::Clip(%677, %678, %679) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%681 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::Mul(%675, %680) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%682 : Float() = onnx::Constant[value={6}]()
%683 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::Div(%681, %682)
%684 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%683, %model.8.cv1.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%685 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%684, %model.8.cv1.bn.weight, %model.8.cv1.bn.bias, %model.8.cv1.bn.running_mean, %model.8.cv1.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%686 : Float() = onnx::Constant[value={3}]()
%687 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Add(%685, %686)
%688 : Tensor = onnx::Constant[value={0}]()
%689 : Tensor = onnx::Constant[value={6}]()
%690 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Clip(%687, %688, %689) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%691 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Mul(%685, %690) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%692 : Float() = onnx::Constant[value={6}]()
%693 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Div(%691, %692)
%694 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::MaxPool[ceil_mode=0, kernel_shape=[5, 5], pads=[2, 2, 2, 2], strides=[1, 1]](%693) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:575:0
%695 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::MaxPool[ceil_mode=0, kernel_shape=[9, 9], pads=[4, 4, 4, 4], strides=[1, 1]](%693) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:575:0
%696 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::MaxPool[ceil_mode=0, kernel_shape=[13, 13], pads=[6, 6, 6, 6], strides=[1, 1]](%693) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:575:0
%697 : Float(1:409600, 1024:400, 20:20, 20:1) = onnx::Concat[axis=1](%693, %694, %695, %696) # /home/ares/文档/yolov5-master-new/models/common.py:79:0
%698 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%697, %model.8.cv2.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%699 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%698, %model.8.cv2.bn.weight, %model.8.cv2.bn.bias, %model.8.cv2.bn.running_mean, %model.8.cv2.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%700 : Float() = onnx::Constant[value={3}]()
%701 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::Add(%699, %700)
%702 : Tensor = onnx::Constant[value={0}]()
%703 : Tensor = onnx::Constant[value={6}]()
%704 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::Clip(%701, %702, %703) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%705 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::Mul(%699, %704) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%706 : Float() = onnx::Constant[value={6}]()
%707 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::Div(%705, %706)
%708 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%707, %model.9.cv1.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%709 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%708, %model.9.cv1.bn.weight, %model.9.cv1.bn.bias, %model.9.cv1.bn.running_mean, %model.9.cv1.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%710 : Float() = onnx::Constant[value={3}]()
%711 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Add(%709, %710)
%712 : Tensor = onnx::Constant[value={0}]()
%713 : Tensor = onnx::Constant[value={6}]()
%714 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Clip(%711, %712, %713) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%715 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Mul(%709, %714) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%716 : Float() = onnx::Constant[value={6}]()
%717 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Div(%715, %716)
%718 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%717, %model.9.m.0.cv1.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%719 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%718, %model.9.m.0.cv1.bn.weight, %model.9.m.0.cv1.bn.bias, %model.9.m.0.cv1.bn.running_mean, %model.9.m.0.cv1.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%720 : Float() = onnx::Constant[value={3}]()
%721 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Add(%719, %720)
%722 : Tensor = onnx::Constant[value={0}]()
%723 : Tensor = onnx::Constant[value={6}]()
%724 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Clip(%721, %722, %723) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%725 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Mul(%719, %724) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%726 : Float() = onnx::Constant[value={6}]()
%727 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Div(%725, %726)
%728 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%727, %model.9.m.0.cv2.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%729 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%728, %model.9.m.0.cv2.bn.weight, %model.9.m.0.cv2.bn.bias, %model.9.m.0.cv2.bn.running_mean, %model.9.m.0.cv2.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%730 : Float() = onnx::Constant[value={3}]()
%731 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Add(%729, %730)
%732 : Tensor = onnx::Constant[value={0}]()
%733 : Tensor = onnx::Constant[value={6}]()
%734 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Clip(%731, %732, %733) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%735 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Mul(%729, %734) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%736 : Float() = onnx::Constant[value={6}]()
%737 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Div(%735, %736)
%738 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%737, %model.9.cv3.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%739 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%707, %model.9.cv2.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%740 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::Concat[axis=1](%738, %739) # /home/ares/文档/yolov5-master-new/models/common.py:65:0
%741 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%740, %model.9.bn.weight, %model.9.bn.bias, %model.9.bn.running_mean, %model.9.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%742 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::LeakyRelu[alpha=0.10000000000000001](%741) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1293:0
%743 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%742, %model.9.cv4.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%744 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%743, %model.9.cv4.bn.weight, %model.9.cv4.bn.bias, %model.9.cv4.bn.running_mean, %model.9.cv4.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%745 : Float() = onnx::Constant[value={3}]()
%746 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::Add(%744, %745)
%747 : Tensor = onnx::Constant[value={0}]()
%748 : Tensor = onnx::Constant[value={6}]()
%749 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::Clip(%746, %747, %748) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%750 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::Mul(%744, %749) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%751 : Float() = onnx::Constant[value={6}]()
%752 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::Div(%750, %751)
%753 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%752, %model.10.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%754 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%753, %model.10.bn.weight, %model.10.bn.bias, %model.10.bn.running_mean, %model.10.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%755 : Float() = onnx::Constant[value={3}]()
%756 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Add(%754, %755)
%757 : Tensor = onnx::Constant[value={0}]()
%758 : Tensor = onnx::Constant[value={6}]()
%759 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Clip(%756, %757, %758) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%760 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Mul(%754, %759) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%761 : Float() = onnx::Constant[value={6}]()
%762 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Div(%760, %761)
%771 : Tensor = onnx::Constant[value=[ CPUFloatType{0} ]]()
%772 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::Resize[coordinate_transformation_mode="asymmetric", cubic_coeff_a=-0.75, mode="nearest", nearest_mode="floor"](%762, %771, %1061) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:3143:0
%773 : Float(1:819200, 512:1600, 40:40, 40:1) = onnx::Concat[axis=1](%772, %673) # /home/ares/文档/yolov5-master-new/models/common.py:99:0
%774 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%773, %model.13.cv1.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%775 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%774, %model.13.cv1.bn.weight, %model.13.cv1.bn.bias, %model.13.cv1.bn.running_mean, %model.13.cv1.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%776 : Float() = onnx::Constant[value={3}]()
%777 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Add(%775, %776)
%778 : Tensor = onnx::Constant[value={0}]()
%779 : Tensor = onnx::Constant[value={6}]()
%780 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Clip(%777, %778, %779) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%781 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Mul(%775, %780) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%782 : Float() = onnx::Constant[value={6}]()
%783 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Div(%781, %782)
%784 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%783, %model.13.m.0.cv1.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%785 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%784, %model.13.m.0.cv1.bn.weight, %model.13.m.0.cv1.bn.bias, %model.13.m.0.cv1.bn.running_mean, %model.13.m.0.cv1.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%786 : Float() = onnx::Constant[value={3}]()
%787 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Add(%785, %786)
%788 : Tensor = onnx::Constant[value={0}]()
%789 : Tensor = onnx::Constant[value={6}]()
%790 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Clip(%787, %788, %789) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%791 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Mul(%785, %790) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%792 : Float() = onnx::Constant[value={6}]()
%793 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Div(%791, %792)
%794 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%793, %model.13.m.0.cv2.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%795 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%794, %model.13.m.0.cv2.bn.weight, %model.13.m.0.cv2.bn.bias, %model.13.m.0.cv2.bn.running_mean, %model.13.m.0.cv2.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%796 : Float() = onnx::Constant[value={3}]()
%797 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Add(%795, %796)
%798 : Tensor = onnx::Constant[value={0}]()
%799 : Tensor = onnx::Constant[value={6}]()
%800 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Clip(%797, %798, %799) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%801 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Mul(%795, %800) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%802 : Float() = onnx::Constant[value={6}]()
%803 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Div(%801, %802)
%804 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%803, %model.13.cv3.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%805 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%773, %model.13.cv2.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%806 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::Concat[axis=1](%804, %805) # /home/ares/文档/yolov5-master-new/models/common.py:65:0
%807 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%806, %model.13.bn.weight, %model.13.bn.bias, %model.13.bn.running_mean, %model.13.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%808 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::LeakyRelu[alpha=0.10000000000000001](%807) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1293:0
%809 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%808, %model.13.cv4.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%810 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%809, %model.13.cv4.bn.weight, %model.13.cv4.bn.bias, %model.13.cv4.bn.running_mean, %model.13.cv4.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%811 : Float() = onnx::Constant[value={3}]()
%812 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::Add(%810, %811)
%813 : Tensor = onnx::Constant[value={0}]()
%814 : Tensor = onnx::Constant[value={6}]()
%815 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::Clip(%812, %813, %814) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%816 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::Mul(%810, %815) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%817 : Float() = onnx::Constant[value={6}]()
%818 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::Div(%816, %817)
%819 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%818, %model.14.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%820 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%819, %model.14.bn.weight, %model.14.bn.bias, %model.14.bn.running_mean, %model.14.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%821 : Float() = onnx::Constant[value={3}]()
%822 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Add(%820, %821)
%823 : Tensor = onnx::Constant[value={0}]()
%824 : Tensor = onnx::Constant[value={6}]()
%825 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Clip(%822, %823, %824) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%826 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Mul(%820, %825) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%827 : Float() = onnx::Constant[value={6}]()
%828 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Div(%826, %827)
%837 : Tensor = onnx::Constant[value=[ CPUFloatType{0} ]]()
%838 : Float(1:819200, 128:6400, 80:80, 80:1) = onnx::Resize[coordinate_transformation_mode="asymmetric", cubic_coeff_a=-0.75, mode="nearest", nearest_mode="floor"](%828, %837, %1066) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:3143:0
%839 : Float(1:1638400, 256:6400, 80:80, 80:1) = onnx::Concat[axis=1](%838, %575) # /home/ares/文档/yolov5-master-new/models/common.py:99:0
%840 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%839, %model.17.cv1.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%841 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%840, %model.17.cv1.bn.weight, %model.17.cv1.bn.bias, %model.17.cv1.bn.running_mean, %model.17.cv1.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%842 : Float() = onnx::Constant[value={3}]()
%843 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Add(%841, %842)
%844 : Tensor = onnx::Constant[value={0}]()
%845 : Tensor = onnx::Constant[value={6}]()
%846 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Clip(%843, %844, %845) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%847 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Mul(%841, %846) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%848 : Float() = onnx::Constant[value={6}]()
%849 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Div(%847, %848)
%850 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%849, %model.17.m.0.cv1.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%851 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%850, %model.17.m.0.cv1.bn.weight, %model.17.m.0.cv1.bn.bias, %model.17.m.0.cv1.bn.running_mean, %model.17.m.0.cv1.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%852 : Float() = onnx::Constant[value={3}]()
%853 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Add(%851, %852)
%854 : Tensor = onnx::Constant[value={0}]()
%855 : Tensor = onnx::Constant[value={6}]()
%856 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Clip(%853, %854, %855) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%857 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Mul(%851, %856) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%858 : Float() = onnx::Constant[value={6}]()
%859 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Div(%857, %858)
%860 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%859, %model.17.m.0.cv2.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%861 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%860, %model.17.m.0.cv2.bn.weight, %model.17.m.0.cv2.bn.bias, %model.17.m.0.cv2.bn.running_mean, %model.17.m.0.cv2.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%862 : Float() = onnx::Constant[value={3}]()
%863 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Add(%861, %862)
%864 : Tensor = onnx::Constant[value={0}]()
%865 : Tensor = onnx::Constant[value={6}]()
%866 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Clip(%863, %864, %865) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%867 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Mul(%861, %866) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%868 : Float() = onnx::Constant[value={6}]()
%869 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Div(%867, %868)
%870 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%869, %model.17.cv3.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%871 : Float(1:409600, 64:6400, 80:80, 80:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%839, %model.17.cv2.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%872 : Float(1:819200, 128:6400, 80:80, 80:1) = onnx::Concat[axis=1](%870, %871) # /home/ares/文档/yolov5-master-new/models/common.py:65:0
%873 : Float(1:819200, 128:6400, 80:80, 80:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%872, %model.17.bn.weight, %model.17.bn.bias, %model.17.bn.running_mean, %model.17.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%874 : Float(1:819200, 128:6400, 80:80, 80:1) = onnx::LeakyRelu[alpha=0.10000000000000001](%873) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1293:0
%875 : Float(1:819200, 128:6400, 80:80, 80:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%874, %model.17.cv4.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%876 : Float(1:819200, 128:6400, 80:80, 80:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%875, %model.17.cv4.bn.weight, %model.17.cv4.bn.bias, %model.17.cv4.bn.running_mean, %model.17.cv4.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%877 : Float() = onnx::Constant[value={3}]()
%878 : Float(1:819200, 128:6400, 80:80, 80:1) = onnx::Add(%876, %877)
%879 : Tensor = onnx::Constant[value={0}]()
%880 : Tensor = onnx::Constant[value={6}]()
%881 : Float(1:819200, 128:6400, 80:80, 80:1) = onnx::Clip(%878, %879, %880) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%882 : Float(1:819200, 128:6400, 80:80, 80:1) = onnx::Mul(%876, %881) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%883 : Float() = onnx::Constant[value={6}]()
%884 : Float(1:819200, 128:6400, 80:80, 80:1) = onnx::Div(%882, %883)
%885 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[2, 2]](%884, %model.18.conv.weight) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%886 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%885, %model.18.bn.weight, %model.18.bn.bias, %model.18.bn.running_mean, %model.18.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%887 : Float() = onnx::Constant[value={3}]()
%888 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Add(%886, %887)
%889 : Tensor = onnx::Constant[value={0}]()
%890 : Tensor = onnx::Constant[value={6}]()
%891 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Clip(%888, %889, %890) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1173:0
%892 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Mul(%886, %891) # /home/ares/文档/yolov5-master-new/utils/activations.py:17:0
%893 : Float() = onnx::Constant[value={6}]()
%894 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Div(%892, %893)
%895 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::Concat[axis=1](%894, %828)
%896 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%895, %model.20.cv1.conv.weight)
%897 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%896, %model.20.cv1.bn.weight, %model.20.cv1.bn.bias, %model.20.cv1.bn.running_mean, %model.20.cv1.bn.running_var)
%898 : Float() = onnx::Constant[value={3}]()
%899 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Add(%897, %898)
%900 : Tensor = onnx::Constant[value={0}]()
%901 : Tensor = onnx::Constant[value={6}]()
%902 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Clip(%899, %900, %901)
%903 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Mul(%897, %902)
%904 : Float() = onnx::Constant[value={6}]()
%905 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Div(%903, %904)
%906 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%905, %model.20.m.0.cv1.conv.weight)
%907 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%906, %model.20.m.0.cv1.bn.weight, %model.20.m.0.cv1.bn.bias, %model.20.m.0.cv1.bn.running_mean, %model.20.m.0.cv1.bn.running_var)
%908 : Float() = onnx::Constant[value={3}]()
%909 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Add(%907, %908)
%910 : Tensor = onnx::Constant[value={0}]()
%911 : Tensor = onnx::Constant[value={6}]()
%912 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Clip(%909, %910, %911)
%913 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Mul(%907, %912)
%914 : Float() = onnx::Constant[value={6}]()
%915 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Div(%913, %914)
%916 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%915, %model.20.m.0.cv2.conv.weight)
%917 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%916, %model.20.m.0.cv2.bn.weight, %model.20.m.0.cv2.bn.bias, %model.20.m.0.cv2.bn.running_mean, %model.20.m.0.cv2.bn.running_var)
%918 : Float() = onnx::Constant[value={3}]()
%919 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Add(%917, %918)
%920 : Tensor = onnx::Constant[value={0}]()
%921 : Tensor = onnx::Constant[value={6}]()
%922 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Clip(%919, %920, %921)
%923 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Mul(%917, %922)
%924 : Float() = onnx::Constant[value={6}]()
%925 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Div(%923, %924)
%926 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%925, %model.20.cv3.weight)
%927 : Float(1:204800, 128:1600, 40:40, 40:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%895, %model.20.cv2.weight)
%928 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::Concat[axis=1](%926, %927)
%929 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%928, %model.20.bn.weight, %model.20.bn.bias, %model.20.bn.running_mean, %model.20.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%930 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::LeakyRelu[alpha=0.10000000000000001](%929) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1293:0
%931 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%930, %model.20.cv4.conv.weight)
%932 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%931, %model.20.cv4.bn.weight, %model.20.cv4.bn.bias, %model.20.cv4.bn.running_mean, %model.20.cv4.bn.running_var)
%933 : Float() = onnx::Constant[value={3}]()
%934 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::Add(%932, %933)
%935 : Tensor = onnx::Constant[value={0}]()
%936 : Tensor = onnx::Constant[value={6}]()
%937 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::Clip(%934, %935, %936)
%938 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::Mul(%932, %937)
%939 : Float() = onnx::Constant[value={6}]()
%940 : Float(1:409600, 256:1600, 40:40, 40:1) = onnx::Div(%938, %939)
%941 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[2, 2]](%940, %model.21.conv.weight)
%942 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%941, %model.21.bn.weight, %model.21.bn.bias, %model.21.bn.running_mean, %model.21.bn.running_var)
%943 : Float() = onnx::Constant[value={3}]()
%944 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Add(%942, %943)
%945 : Tensor = onnx::Constant[value={0}]()
%946 : Tensor = onnx::Constant[value={6}]()
%947 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Clip(%944, %945, %946)
%948 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Mul(%942, %947)
%949 : Float() = onnx::Constant[value={6}]()
%950 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Div(%948, %949)
%951 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::Concat[axis=1](%950, %762)
%952 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%951, %model.23.cv1.conv.weight)
%953 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%952, %model.23.cv1.bn.weight, %model.23.cv1.bn.bias, %model.23.cv1.bn.running_mean, %model.23.cv1.bn.running_var)
%954 : Float() = onnx::Constant[value={3}]()
%955 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Add(%953, %954)
%956 : Tensor = onnx::Constant[value={0}]()
%957 : Tensor = onnx::Constant[value={6}]()
%958 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Clip(%955, %956, %957)
%959 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Mul(%953, %958)
%960 : Float() = onnx::Constant[value={6}]()
%961 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Div(%959, %960)
%962 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%961, %model.23.m.0.cv1.conv.weight)
%963 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%962, %model.23.m.0.cv1.bn.weight, %model.23.m.0.cv1.bn.bias, %model.23.m.0.cv1.bn.running_mean, %model.23.m.0.cv1.bn.running_var)
%964 : Float() = onnx::Constant[value={3}]()
%965 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Add(%963, %964)
%966 : Tensor = onnx::Constant[value={0}]()
%967 : Tensor = onnx::Constant[value={6}]()
%968 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Clip(%965, %966, %967)
%969 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Mul(%963, %968)
%970 : Float() = onnx::Constant[value={6}]()
%971 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Div(%969, %970)
%972 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[3, 3], pads=[1, 1, 1, 1], strides=[1, 1]](%971, %model.23.m.0.cv2.conv.weight)
%973 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%972, %model.23.m.0.cv2.bn.weight, %model.23.m.0.cv2.bn.bias, %model.23.m.0.cv2.bn.running_mean, %model.23.m.0.cv2.bn.running_var)
%974 : Float() = onnx::Constant[value={3}]()
%975 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Add(%973, %974)
%976 : Tensor = onnx::Constant[value={0}]()
%977 : Tensor = onnx::Constant[value={6}]()
%978 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Clip(%975, %976, %977)
%979 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Mul(%973, %978)
%980 : Float() = onnx::Constant[value={6}]()
%981 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Div(%979, %980)
%982 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%981, %model.23.cv3.weight)
%983 : Float(1:102400, 256:400, 20:20, 20:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%951, %model.23.cv2.weight)
%984 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::Concat[axis=1](%982, %983)
%985 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%984, %model.23.bn.weight, %model.23.bn.bias, %model.23.bn.running_mean, %model.23.bn.running_var) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:2014:0
%986 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::LeakyRelu[alpha=0.10000000000000001](%985) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/functional.py:1293:0
%987 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%986, %model.23.cv4.conv.weight)
%988 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::BatchNormalization[epsilon=0.001, momentum=0.96999999999999997](%987, %model.23.cv4.bn.weight, %model.23.cv4.bn.bias, %model.23.cv4.bn.running_mean, %model.23.cv4.bn.running_var)
%989 : Float() = onnx::Constant[value={3}]()
%990 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::Add(%988, %989)
%991 : Tensor = onnx::Constant[value={0}]()
%992 : Tensor = onnx::Constant[value={6}]()
%993 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::Clip(%990, %991, %992)
%994 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::Mul(%988, %993)
%995 : Float() = onnx::Constant[value={6}]()
%996 : Float(1:204800, 512:400, 20:20, 20:1) = onnx::Div(%994, %995)
%997 : Float(1:134400, 21:6400, 80:80, 80:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%884, %model.24.m.0.weight, %model.24.m.0.bias) # /home/ares/anaconda3/envs/yolo/lib/python3.8/site-packages/torch/nn/modules/conv.py:415:0
%998 : Tensor = onnx::Shape(%997)
%999 : Tensor = onnx::Constant[value={0}]()
%1000 : Long() = onnx::Gather[axis=0](%998, %999
%1001 : Tensor = onnx::Shape(%997)
%1002 : Tensor = onnx::Constant[value={2}]()
%1003 : Long() = onnx::Gather[axis=0](%1001, %1002)
%1004 : Tensor = onnx::Shape(%997)
%1005 : Tensor = onnx::Constant[value={3}]()
%1006 : Long() = onnx::Gather[axis=0](%1004, %1005)
%1009 : Tensor = onnx::Unsqueeze[axes=[0]](%1000)
%1012 : Tensor = onnx::Unsqueeze[axes=[0]](%1003)
%1013 : Tensor = onnx::Unsqueeze[axes=[0]](%1006)
%1014 : Tensor = onnx::Concat[axis=0](%1009, %1067, %1068, %1012, %1013)
%1015 : Float(1:134400, 3:44800, 7:6400, 80:80, 80:1) = onnx::Reshape(%997, %1014)
%classes : Float(1:134400, 3:44800, 80:560, 80:7, 7:1) = onnx::Transpose[perm=[0, 1, 3, 4, 2]](%1015)
%1017 : Float(1:33600, 21:1600, 40:40, 40:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%940, %model.24.m.1.weight, %model.24.m.1.bias)
%1018 : Tensor = onnx::Shape(%1017)
%1019 : Tensor = onnx::Constant[value={0}]()
%1020 : Long() = onnx::Gather[axis=0](%1018, %1019)
%1021 : Tensor = onnx::Shape(%1017)
%1022 : Tensor = onnx::Constant[value={2}]()
%1023 : Long() = onnx::Gather[axis=0](%1021, %1022)
%1024 : Tensor = onnx::Shape(%1017)
%1025 : Tensor = onnx::Constant[value={3}]()
%1026 : Long() = onnx::Gather[axis=0](%1024, %1025)
%1029 : Tensor = onnx::Unsqueeze[axes=[0]](%1020)
%1032 : Tensor = onnx::Unsqueeze[axes=[0]](%1023)
%1033 : Tensor = onnx::Unsqueeze[axes=[0]](%1026)
%1034 : Tensor = onnx::Concat[axis=0](%1029, %1069, %1070, %1032, %1033)
%1035 : Float(1:33600, 3:11200, 7:1600, 40:40, 40:1) = onnx::Reshape(%1017, %1034)
%boxes : Float(1:33600, 3:11200, 40:280, 40:7, 7:1) = onnx::Transpose[perm=[0, 1, 3, 4, 2]](%1035)
%1037 : Float(1:8400, 21:400, 20:20, 20:1) = onnx::Conv[dilations=[1, 1], group=1, kernel_shape=[1, 1], pads=[0, 0, 0, 0], strides=[1, 1]](%996, %model.24.m.2.weight, %model.24.m.2.bias)
%1038 : Tensor = onnx::Shape(%1037)
%1039 : Tensor = onnx::Constant[value={0}]()
%1040 : Long() = onnx::Gather[axis=0](%1038, %1039)
%1041 : Tensor = onnx::Shape(%1037)
%1042 : Tensor = onnx::Constant[value={2}]()
%1043 : Long() = onnx::Gather[axis=0](%1041, %1042)
%1044 : Tensor = onnx::Shape(%1037)
%1045 : Tensor = onnx::Constant[value={3}]()
%1046 : Long() = onnx::Gather[axis=0](%1044, %1045)
%1049 : Tensor = onnx::Unsqueeze[axes=[0]](%1040)
%1052 : Tensor = onnx::Unsqueeze[axes=[0]](%1043)
%1053 : Tensor = onnx::Unsqueeze[axes=[0]](%1046)
%1054 : Tensor = onnx::Concat[axis=0](%1049, %1071, %1072, %1052, %1053)
%1055 : Float(1:8400, 3:2800, 7:400, 20:20, 20:1) = onnx::Reshape(%1037, %1054)
%1056 : Float(1:8400, 3:2800, 20:140, 20:7, 7:1) = onnx::Transpose[perm=[0, 1, 3, 4, 2]](%1055)
return (%classes, %boxes, %1056)
On jetson nano, run detectnet
detectnet --model=best.onnx --labels=models/labels.txt --threshold=0.8 --input-blob=input --output-cvg=classes --output-bbox=boxes /dev/video0
**************
**After tensorrt**
**fail report**
**************
RingBuffer -- allocated 4 buffers (2764800 bytes each, 11059200 bytes total)
[TRT] engine.cpp (986) - Cuda Error in executeInternal: 719 (unspecified launch failure)
[TRT] FAILED_EXECUTION: std::exception
[TRT] failed to execute TensorRT context on device GPU
Traceback (most recent call last):
File "/usr/local/bin/detectnet.py", line 63, in <module>
detections = net.Detect(img, overlay=opt.overlay)
Exception: jetson.inference -- detectNet.Detect() encountered an error classifying the image
[cuda] unspecified launch failure (error 719) (hex 0x2CF)
[cuda] /home/ares/文档/jetson-inference/c/detectNet.cpp:68
[cuda] unspecified launch failure (error 719) (hex 0x2CF)
[cuda] /home/ares/文档/jetson-inference/c/detectNet.cpp:76
[TRT] ../rtExt/cuda/cudaFusedConvActRunner.cpp (90) - Cuda Error in destroyFilterTexture: 719 (unspecified launch failure)
[TRT] INTERNAL_ERROR: std::exception
[TRT] ../rtSafe/safeRuntime.cpp (32) - Cuda Error in free: 719 (unspecified launch failure)
terminate called after throwing an instance of 'nvinfer1::CudaError'
what(): std::exception
已放弃 (核心已转储)