请问这个您的yolo.py可以用旷视训练好的yolox_s的直接做预测推理吗？

[Maxwell2016LeChouchou]

请问这个您的yolo.py和predict.py可以用旷视训练好的yolox_s的直接做预测推理吗？ 我看您的predict.py有可视化heatmap，就想用我在旷视训练好的yolox_s（自己标注的训练集）在您的代码上直接做heatmap可视化。 但是报一大堆backbone的错误，主要是这些：

(open-mmlab) C:\Maxwell_W\detection\yolox-pytorch>python predict.py img/clean_AI_00822.png
Traceback (most recent call last):
  File "predict.py", line 14, in <module>
    yolo = YOLO()
  File "C:\Maxwell_W\detection\yolox-pytorch\yolo.py", line 85, in __init__
    self.generate()
  File "C:\Maxwell_W\detection\yolox-pytorch\yolo.py", line 95, in generate
    self.net.load_state_dict(torch.load(self.model_path, map_location=device))
  File "C:\Users\potentialmotors\miniconda3\envs\open-mmlab\lib\site-packages\torch\nn\modules\module.py", line 1671, in load_state_dict
    raise RuntimeError('Error(s) in loading state_dict for {}:\n\t{}'.format(
RuntimeError: Error(s) in loading state_dict for YoloBody:
        Missing key(s) in state_dict: "backbone.backbone.stem.conv.conv.weight", "[backbone.backbone.stem.conv.bn](http://backbone.backbone.stem.conv.bn/).weight", "[backbone.backbone.stem.conv.bn](http://backbone.backbone.stem.conv.bn/).bias", "backbone.backbone.stem.conv.bn.running_mean",
"""""""
这里报一大堆的backbone nn的错误 
"""""""
        Unexpected key(s) in state_dict: "start_epoch", "model", "optimizer", "best_ap", "curr_ap".

[illrayy]

模型的权重是通过字典加载的：

pretrained_dict = torch.load(model_path, map_location = device)
pretrained_dict = {k: v for k, v in pretrained_dict.items() if np.shape(model_dict[k]) == np.shape(v)}

官方仓库训练的模型和这个仓库的key不同，你可以对一下两个权重的shape，看看这个仓库的key是什么样的，然后自己改一下官方的key