Pytorch_Retinaface python3 detect.py --save_model err

[HBProlove]

Env

• GPU: Xavier NX
• OS,: Ubuntu18.04.
• NVIDIA Jetson Xavier NX (Developer Kit Version)
  • Jetpack 4.5.1 [L4T 32.5.1]
  • NV Power Mode: MODE_15W_6CORE - Type: 2
  • jetson_stats.service: active
• Libraries:
  • CUDA: 10.2.89
  • cuDNN: 8.0.0.180
  • TensorRT: 7.1.3.0
  • Visionworks: 1.6.0.501
  • OpenCV: 4.1.1 compiled CUDA: NO
  • VPI: ii libnvvpi1 1.0.15 arm64 NVIDIA Vision Programming Interface library
  • Vulkan: 1.2.70

About this repo

• which branch/tag/commit are you using?
  master
  commit beda45c3665f48a20dc2fba5a8e91032100c50c4 (HEAD -> master, origin/master, origin/HEAD)
  Merge: 00563fe b984b4b
  Author: Wang Xinyu [email protected]
  Date: Thu May 14 20:59:07 2020 +0800
  Merge pull request add mobilenetv3 small #1 from biubug6/master
  Update README.md
• which model? yolov5, retinaface?
  retinaface

Your problem

a411@a411-nx:~/RetinaFace_trt_proj/Pytorch_Retinaface$ python3 detect.py --save_model
Loading pretrained model from ./weights/Resnet50_Final.pth
remove prefix 'module.'
Missing keys:456
Unused checkpoint keys:267
Used keys:0
Traceback (most recent call last):
File "detect.py", line 76, in
net = load_model(net, args.trained_model, args.cpu)
File "detect.py", line 62, in load_model
check_keys(model, pretrained_dict)
File "detect.py", line 40, in check_keys
assert len(used_pretrained_keys) > 0, 'load NONE from pretrained checkpoint'
AssertionError: load NONE from pretrained checkpoint

[wang-xinyu]

Did you follow here strictly https://github.com/wang-xinyu/tensorrtx/tree/master/retinaface#run?

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.