pytorch重新加载参数的代码，具体步骤

#这个参数主要是用来设置是否从断点处继续训练
    if args.resume:
        #if isfile(args.resume):
 “

isfile(args.resume):里面的（）放入自己的最好模型的具体位置

print("=> loading checkpoint '{}'".format(放入自己的具体路径）

checkpoint = torch.load(具体位置)

logger = Logger(join(args.checkpoint, （具体位置））

终端输入形式：CUDA_VISIBLE_DEVICES=0 python example/mpii.py -a hg --stacks 4 --blocks 1 --resume checkpoint/mpii/hg4/model_best.pth.tar -j 4

 

 

 

”

 

 

 

 if isfile('/media/z/CC/zxl/pytorch-pose/checkpoint/mpii/hg8/model_best.pth.tar'): print("=> loading checkpoint '{}'".format('/media/z/CC/zxl/pytorch-pose/checkpoint/mpii/hg8/model_best.pth.tar')) # checkpoint = torch.load(args.resume)#是用来导入已训练好的模型 checkpoint = torch.load('/media/z/CC/zxl/pytorch-pose/checkpoint/mpii/hg8/model_best.pth.tar') args.start_epoch = checkpoint['epoch'] best_acc = checkpoint['best_acc'] model.load_state_dict(checkpoint['state_dict'])#完成导入模型的参数初始化model这个网络的过程 #model.load_state_dict(checkpoint['/media/z/CC/zxl/pytorch-pose/checkpoint /mpii/hg8/model_best.pth.tar']) optimizer.load_state_dict(checkpoint['optimizer']) print("=> loaded checkpoint '{}' (epoch {})" .format(args.resume, checkpoint['epoch'])) # logger = Logger(join(args.checkpoint, 'log.txt'), title=title, resume=True) logger = Logger(join(args.checkpoint, '/media/z/CC/zxl/pytorch-pose/checkpoint/mpii/hg8/log.txt'), title=title, resume=True) else: print("=> no checkpoint found at '{}'".format(args.resume)) else: