WZMIAOMIAO/deep-learning-for-image-processing · error · ValueError

In training mode, targets should be passed

Error message

In training mode, targets should be passed

What it means

FasterRCNN.forward requires a `targets` list when the module is in training mode (self.training == True); calling it with targets=None raises this ValueError. During training the model must compute losses against ground-truth boxes/labels, which is impossible without targets.

Source

Thrown at pytorch_object_detection/faster_rcnn/network_files/faster_rcnn_framework.py:60

        return detections

    def forward(self, images, targets=None):
        # type: (List[Tensor], Optional[List[Dict[str, Tensor]]]) -> Tuple[Dict[str, Tensor], List[Dict[str, Tensor]]]
        """
        Arguments:
            images (list[Tensor]): images to be processed
            targets (list[Dict[Tensor]]): ground-truth boxes present in the image (optional)

        Returns:
            result (list[BoxList] or dict[Tensor]): the output from the model.
                During training, it returns a dict[Tensor] which contains the losses.
                During testing, it returns list[BoxList] contains additional fields
                like `scores`, `labels` and `mask` (for Mask R-CNN models).

        """
        if self.training and targets is None:
            raise ValueError("In training mode, targets should be passed")

        if self.training:
            assert targets is not None
            for target in targets:         # 进一步判断传入的target的boxes参数是否符合规定
                boxes = target["boxes"]
                if isinstance(boxes, torch.Tensor):
                    if len(boxes.shape) != 2 or boxes.shape[-1] != 4:
                        raise ValueError("Expected target boxes to be a tensor"
                                         "of shape [N, 4], got {:}.".format(
                                          boxes.shape))
                else:
                    raise ValueError("Expected target boxes to be of type "
                                     "Tensor, got {:}.".format(type(boxes)))

        original_image_sizes = torch.jit.annotate(List[Tuple[int, int]], [])
        for img in images:
            val = img.shape[-2:]
            assert len(val) == 2  # 防止输入的是个一维向量

View on GitHub (pinned to 1ec3fe6f37)

Solutions

1. Ensure the model is in eval mode (model.eval()) if you only want predictions.
2. In the training loop, unpack and pass targets: loss_dict = model(images, targets).
3. Verify the dataset/collate_fn returns (image, target) pairs and targets is a list of dicts with 'boxes' and 'labels'.
4. Wrap targets=None cases: skip training steps where targets are missing.

Example fix

# before
for images in train_loader:
    loss_dict = model(images)
# after
for images, targets in train_loader:
    loss_dict = model(images, targets)

if model.training:
    assert targets is not None and isinstance(targets, list) and len(targets) == len(images), "targets required in training mode"

def has_targets(batch) -> bool:
    images, targets = batch
    return targets is not None and len(targets) == len(images) and all('boxes' in t and 'labels' in t for t in targets)

try:
    loss_dict = model(images, targets)
except ValueError as e:
    if 'targets should be passed' in str(e):
        raise RuntimeError('Training loop must supply targets; use model.eval() for inference') from e
    raise

When it happens

Trigger: Calling model(images) without the second argument (or passing None) after model.train(), e.g. forgetting to unpack both outputs of the dataloader: for images in loader instead of for images, targets in loader.

Common situations: Copy-pasting inference code into a training loop, forgetting model.eval() before inference-only calls, or a dataloader yielding only images because the dataset doesn't return targets.

Related errors

• Expected target boxes to be a tensorof shape [N, 4], got {:}
• Expected target boxes to be of type Tensor, got {:}.
• num_classes should be None when box_predictor is specified
• In training mode, targets should be passed
• In training mode, targets should be passed

AI-assisted analysis of WZMIAOMIAO/deep-learning-for-image-processing@1ec3fe6f37 (2026-08-30). Data as JSON: /api/errors/389042055d5ddb09. Report an issue: GitHub.