Successfully convert .pt to weights file

Successfully implement .pt to .weights file

Recently, I used the pytorch framework to train pictures, and generated a .pt file. When I
finally used it, I used the .weights file. I tried several codes. Finally, I succeeded in using this. The .pt to .weights is available.

Create a new pt2weights.py file

from models import *  //注意找到工程文件中有没有 models.py文件 例如下面Darknet save_weights函数都在models.py文件里定义好了  直接调用就好  注意：新建的pt2weights.py与models.py在同一目录下
model=Darknet("cfg/yolov3.cfg") //训练时候用的cfg 改成自己.cfg的地址 
#load_darknet_weights(model,"weights/latest.pt")
#save_weights(model,path='weights/latest.weights',cutoff=-1)
checkpoint = torch.load("weights//200轮修改anchorbox/best.pt",map_location='cpu')//导入训练好的.pt文件 改成自己的.pt文件地址
model.load_state_dict(checkpoint['model'])
save_weights(model,path='weights/200轮修改anchorbox/200_xiu_best.weights',cutoff=-1)//生成.weights文件 改成自己的地址