RuntimeError: Found dtype Double but expected Float with Faster RCNN - Pytorch

Question

I am new one for Deep Learning, when I use Faster RCNN on Kaggle to detect my own dataset, I face to this error which occurs on line 18 losses.backward(), and I have tried image.to(torch.float32) and targets.to(torch.float32), it wasn't work, I have no idea on it. Here is my code, please help and thanks alot.

loss_hist = Averager()
itr = 1

for epoch in range(num_epochs):
    loss_hist.reset()
    
    for images, targets, image_ids in train_data_loader:
        
        images = list(image.to(device) for image in images)
        targets = [{k: v.to(device) for k, v in t.items()} for t in targets]
        
        loss_dict = frcnn(images, targets)
        losses = sum(loss for loss in loss_dict.values())
        loss_value = losses.item()

        loss_hist.send(loss_value)

        optimizer.zero_grad()
        losses.backward()
        optimizer.step()

        if itr % 50 == 0:
            print(f"Iteration #{itr} loss: {loss_value}")

        itr += 1
    
    # update the learning rate
    if lr_scheduler is not None:
                lr_scheduler.step()

    print(f"Epoch #{epoch} loss: {loss_hist.value}")   

and the error is here

---------------------------------------------------------------------------
RuntimeError                              Traceback (most recent call last)
/tmp/ipykernel_1418/2816814501.py in <module>
     17 
     18         optimizer.zero_grad()
---> 19         losses.backward()
     20         optimizer.step()
     21 

/opt/conda/lib/python3.7/site-packages/torch/_tensor.py in backward(self, gradient, retain_graph, create_graph, inputs)
    253                 create_graph=create_graph,
    254                 inputs=inputs)
--> 255         torch.autograd.backward(self, gradient, retain_graph, create_graph, inputs=inputs)
    256 
    257     def register_hook(self, hook):

/opt/conda/lib/python3.7/site-packages/torch/autograd/__init__.py in backward(tensors, grad_tensors, retain_graph, create_graph, grad_variables, inputs)
    147     Variable._execution_engine.run_backward(
    148         tensors, grad_tensors_, retain_graph, create_graph, inputs,
--> 149         allow_unreachable=True, accumulate_grad=True)  # allow_unreachable flag
    150 
    151 

RuntimeError: Found dtype Double but expected Float

Here is my input

print(images[1])
tensor([[[0.8353, 0.8392, 0.8431,  ..., 0.1569, 0.1608, 0.1569],
         [0.8392, 0.8431, 0.8431,  ..., 0.1686, 0.1686, 0.1647],
         [0.8471, 0.8510, 0.8510,  ..., 0.1804, 0.1765, 0.1686],
         ...,
         [0.3804, 0.3843, 0.3922,  ..., 0.9137, 0.9137, 0.9137],
         [0.3882, 0.3922, 0.4039,  ..., 0.9098, 0.9098, 0.9176],
         [0.4000, 0.4078, 0.4118,  ..., 0.9020, 0.9020, 0.9059]]],
       device='cuda:0')

print(targets[1])
{'boxes': tensor([[2668.2204, 1033.8983, 2932.6270, 1247.4576]], device='cuda:0',
       dtype=torch.float64), 'labels': tensor([2], device='cuda:0'), 'image_id': tensor([52], device='cuda:0'), 'area': tensor([56466.4922], device='cuda:0'), 'iscrowd': tensor([0], device='cuda:0')}

---

Here is my loss funtion

class Averager:
    def __init__(self):
        self.current_total = 0.0
        self.iterations = 0.0

    def send(self, value):
        self.current_total += value
        self.iterations += 1

    @property
    def value(self):
        if self.iterations == 0:
            return 0
        else:
            return 1.0 * self.current_total / self.iterations

    def reset(self):
        self.current_total = 0.0
        self.iterations = 0.0