RuntimeError: Expected u.is_cuda() to be true, but got false. (Could this error message be improved? If so, please report an enhancement request to PyTorch.)`

[sekioo]

Traceback (most recent call last): File "/home/renzecheng/gcx/ultralytics-main/mbyolo_train.py", line 44, in <module> "train": YOLO(model_conf).train(**args), File "/home/renzecheng/gcx/ultralytics-main/ultralytics/models/yolo/model.py", line 23, in __init__ super().__init__(model=model, task=task, verbose=verbose) File "/home/renzecheng/gcx/ultralytics-main/ultralytics/engine/model.py", line 143, in __init__ self._new(model, task=task, verbose=verbose) File "/home/renzecheng/gcx/ultralytics-main/ultralytics/engine/model.py", line 251, in _new self.model = (model or self._smart_load("model"))(cfg_dict, verbose=verbose and RANK == -1) # build model File "/home/renzecheng/gcx/ultralytics-main/ultralytics/nn/tasks.py", line 342, in __init__ m.stride = torch.tensor([s / x.shape[-2] for x in _forward(torch.zeros(1, ch, s, s))]) # forward File "/home/renzecheng/gcx/ultralytics-main/ultralytics/nn/tasks.py", line 340, in _forward return self.forward(x)[0] if isinstance(m, (Segment, Pose, OBB)) else self.forward(x) File "/home/renzecheng/gcx/ultralytics-main/ultralytics/nn/tasks.py", line 118, in forward return self.predict(x, *args, **kwargs) File "/home/renzecheng/gcx/ultralytics-main/ultralytics/nn/tasks.py", line 136, in predict return self._predict_once(x, profile, visualize, embed) File "/home/renzecheng/gcx/ultralytics-main/ultralytics/nn/tasks.py", line 157, in _predict_once x = m(x) # run File "/home/renzecheng/anaconda3/envs/mambaYolo-11/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1518, in _wrapped_call_impl return self._call_impl(*args, **kwargs) File "/home/renzecheng/anaconda3/envs/mambaYolo-11/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1527, in _call_impl return forward_call(*args, **kwargs) File "/home/renzecheng/gcx/ultralytics-main/ultralytics/nn/modules/mamba_yolo.py", line 382, in forward x = input + self.drop_path(self.op(self.norm(X1))) File "/home/renzecheng/anaconda3/envs/mambaYolo-11/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1518, in _wrapped_call_impl return self._call_impl(*args, **kwargs) File "/home/renzecheng/anaconda3/envs/mambaYolo-11/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1527, in _call_impl return forward_call(*args, **kwargs) File "/home/renzecheng/gcx/ultralytics-main/ultralytics/nn/modules/mamba_yolo.py", line 186, in forward y = self.forward_core(x, channel_first=(self.d_conv > 1)) File "/home/renzecheng/gcx/ultralytics-main/ultralytics/nn/modules/mamba_yolo.py", line 165, in forward_corev2 x = cross_selective_scan( File "/home/renzecheng/gcx/ultralytics-main/ultralytics/nn/modules/common_utils_mbyolo.py", line 192, in cross_selective_scan ys: torch.Tensor = selective_scan( File "/home/renzecheng/gcx/ultralytics-main/ultralytics/nn/modules/common_utils_mbyolo.py", line 168, in selective_scan return SelectiveScan.apply(u, delta, A, B, C, D, delta_bias, delta_softplus, nrows, backnrows, ssoflex) File "/home/renzecheng/anaconda3/envs/mambaYolo-11/lib/python3.10/site-packages/torch/autograd/function.py", line 539, in apply return super().apply(*args, **kwargs) # type: ignore[misc] File "/home/renzecheng/anaconda3/envs/mambaYolo-11/lib/python3.10/site-packages/torch/cuda/amp/autocast_mode.py", line 113, in decorate_fwd return fwd(*args, **kwargs) File "/home/renzecheng/gcx/ultralytics-main/ultralytics/nn/modules/common_utils_mbyolo.py", line 125, in forward out, x, *rest = selective_scan_cuda_core.fwd(u, delta, A, B, C, D, delta_bias, delta_softplus, 1) RuntimeError: Expected u.is_cuda() to be true, but got false. (Could this error message be improved? If so, please report an enhancement request to PyTorch.)

[sekioo]

I tried to carry your all of Mamba modules in YOLOV8 to YOLOV11, but this issue appeared, please help me solve it,thanks

[lishaobo-GitHub]

In yolov9, you need edit class DetectionModel(BaseModel): for YOLOv8 and YOLOv11 is same method
class DetectionModel(BaseModel):
"""YOLOv8 detection model."""

def __init__(self, cfg="yolov8n.yaml", ch=3, nc=None, verbose=True):  # model, input channels, number of classes
    """Initialize the YOLOv8 detection model with the given config and parameters."""
    super().__init__()
    self.yaml = cfg if isinstance(cfg, dict) else yaml_model_load(cfg)  # cfg dict

    # Define model
    ch = self.yaml["ch"] = self.yaml.get("ch", ch)  # input channels
    if nc and nc != self.yaml["nc"]:
        LOGGER.info(f"Overriding model.yaml nc={self.yaml['nc']} with nc={nc}")
        self.yaml["nc"] = nc  # override YAML value
    self.model, self.save = parse_model(deepcopy(self.yaml), ch=ch, verbose=verbose)  # model, savelist
    self.names = {i: f"{i}" for i in range(self.yaml["nc"])}  # default names dict
    self.inplace = self.yaml.get("inplace", True)

    # Build strides
    m = self.model[-1]  # Detect()
    if isinstance(m, Detect):  # includes all Detect subclasses like Segment, Pose, OBB, WorldDetect
        s = 256  # 2x min stride
        m.inplace = self.inplace
        forward = lambda x: self.forward(x)[0] if isinstance(m, (Segment, Pose, OBB)) else self.forward(x)
        try: #there add exception
            m.stride = torch.tensor([s / x.shape[-2] for x in forward(torch.zeros(1, ch, s, s))])  # forward
        except Exception as e:
            if 'Not implemented on the CPU' in str(
                    e) or 'Input type (torch.FloatTensor) and weight type (torch.cuda.FloatTensor)' in str(
                e) or 'CUDA tensor' in str(e) or 'is_cuda()' in str(e) or "Cannot find backend for cpu" in str(e):
                self.model.to(torch.device('cuda'))
                m.stride = torch.tensor([s / x.shape[-2] for x in
                                         forward(torch.zeros(2, ch, s, s).to(torch.device('cuda')))])  # forward
            else:
                raise e
        self.stride = m.stride
        m.bias_init()  # only run once
    else:
        self.stride = torch.Tensor([32])  # default stride for i.e. RTDETR

    # Init weights, biases
    initialize_weights(self)
    if verbose:
        self.info()
        LOGGER.info("")