(33)assert all(map(lambda i: i.is_cuda, inputs)) AssertionError

问题描述：在跑Faster R-CNN时，使用单GPU训练，计算mAP时，报错：

logs/logs_1/ep150-loss1.017-val_loss1.106.pth model, anchors, and classes loaded.
  0%|          | 0/4952 [00:00
    frcnn.get_map_txt(image_id, image, class_names, map_out_path)
  File "/home/PycharmProjects/compare-programs/faster-rcnn-pytorch-master/frcnn.py", line 278, in get_map_txt
    roi_cls_locs, roi_scores, rois, _ = self.net(images)
  File "/home/.conda/envs/ultralytics_yolov3/lib/python3.7/site-packages/torch/nn/modules/module.py", line 722, in _call_impl
    result = self.forward(*input, **kwargs)
  File "/home/.conda/envs/ultralytics_yolov3/lib/python3.7/site-packages/torch/nn/parallel/data_parallel.py", line 156, in forward
    return self.gather(outputs, self.output_device)
  File "/home/.conda/envs/ultralytics_yolov3/lib/python3.7/site-packages/torch/nn/parallel/data_parallel.py", line 168, in gather
    return gather(outputs, output_device, dim=self.dim)
  File "/home/.conda/envs/ultralytics_yolov3/lib/python3.7/site-packages/torch/nn/parallel/scatter_gather.py", line 68, in gather
    res = gather_map(outputs)
  File "/home/.conda/envs/ultralytics_yolov3/lib/python3.7/site-packages/torch/nn/parallel/scatter_gather.py", line 63, in gather_map
    return type(out)(map(gather_map, zip(*outputs)))
  File "/home/.conda/envs/ultralytics_yolov3/lib/python3.7/site-packages/torch/nn/parallel/scatter_gather.py", line 55, in gather_map
    return Gather.apply(target_device, dim, *outputs)
  File "/home/.conda/envs/ultralytics_yolov3/lib/python3.7/site-packages/torch/nn/parallel/_functions.py", line 54, in forward
    assert all(map(lambda i: i.is_cuda, inputs))
AssertionError

问题分析：是由于电脑上有两个显卡，但是代码使用的单GPU，因此在未指定代码用哪个GPU时，它不知道使用哪个。

问题解决：定位问题到frcnn.py的278行，该部分代码是生成模型的代码块，原始代码为：

#---------------------------------------------------#
    #   载入模型
    #---------------------------------------------------#
    def generate(self):
        #-------------------------------#
        #   载入模型与权值
        #-------------------------------#
        self.net    = FasterRCNN(self.num_classes, "predict", anchor_scales = self.anchors_size, backbone = self.backbone)
        device      = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
        self.net.load_state_dict(torch.load(self.model_path, map_location=device))
        self.net    = self.net.eval()
        print('{} model, anchors, and classes loaded.'.format(self.model_path))
        
        if self.cuda:
            self.net = nn.DataParallel(self.net)
            self.net = self.net.cuda()

为模型指定使用哪个GPU，将最后两行修改为：

if self.cuda:
   self.net = nn.DataParallel(self.net, device_ids=[0])
   self.net = self.net.cuda()

问题解决。

参考：

assert all(map(lambda i: i.is_cuda, inputs)) AssertionError_学无止境、积少成多、厚积薄发-CSDN博客