paddleYOLOV5 : The meta data must be valid when call the mutable data function.

[zhangxp12]

问题确认 Search before asking

• [x] 我已经查询历史issue，没有发现相似的bug。I have searched the issues and found no similar bug report.

Bug组件 Bug Component

No response

Bug描述 Describe the Bug

(PaddleYOLO) D:\project\GitHub\PaddleYOLO\PaddleYOLO>python tools/train.py -c configs/yolov5/yolov5_x_300e_coco.yml --eval loading annotations into memory... Done (t=0.01s) creating index... index created! [05/06 13:56:41] ppdet.data.source.coco INFO: Load [230 samples valid, 0 samples invalid] in file dataset/coco2/annotations/train.json. W0506 13:56:41.844275 10804 gpu_resources.cc:61] Please NOTE: device: 0, GPU Compute Capability: 6.1, Driver API Version: 12.1, Runtime API Version: 10.2 W0506 13:56:41.851253 10804 gpu_resources.cc:91] device: 0, cuDNN Version: 7.6. [05/06 13:56:45] ppdet.engine INFO: Epoch: [0] [ 0/230] eta: 17:05:51 lr: 0.000000 loss_box: 0.117473 loss_obj: 2.690734 loss_cls: 0.000000 loss: 2.808207 batch_cost: 1.3381 data_cost: 0.0000 ips: 0.7473 images/s Traceback (most recent call last): File "D:\project\GitHub\PaddleYOLO\PaddleYOLO\tools\train.py", line 188, in main() File "D:\project\GitHub\PaddleYOLO\PaddleYOLO\tools\train.py", line 184, in main run(FLAGS, cfg) File "D:\project\GitHub\PaddleYOLO\PaddleYOLO\tools\train.py", line 137, in run trainer.train(FLAGS.eval) File "D:\project\GitHub\PaddleYOLO\PaddleYOLO\ppdet\engine\trainer.py", line 433, in train outputs = model(data) File "D:\Program Files (x86)\python_package\PaddleYOLO\lib\site-packages\paddle\fluid\dygraph\layers.py", line 930, in call return self._dygraph_call_func(*inputs, kwargs) File "D:\Program Files (x86)\python_package\PaddleYOLO\lib\site-packages\paddle\fluid\dygraph\layers.py", line 915, in _dygraph_call_func outputs = self.forward(*inputs, *kwargs) File "D:\project\GitHub\PaddleYOLO\PaddleYOLO\ppdet\modeling\architectures\meta_arch.py", line 59, in forward out = self.get_loss() File "D:\project\GitHub\PaddleYOLO\PaddleYOLO\ppdet\modeling\architectures\yolov5.py", line 95, in get_loss return self._forward() File "D:\project\GitHub\PaddleYOLO\PaddleYOLO\ppdet\modeling\architectures\yolov5.py", line 77, in _forward yolo_losses = self.yolo_head(neck_feats, self.inputs) File "D:\Program Files (x86)\python_package\PaddleYOLO\lib\site-packages\paddle\fluid\dygraph\layers.py", line 930, in call return self._dygraph_call_func(inputs, kwargs) File "D:\Program Files (x86)\python_package\PaddleYOLO\lib\site-packages\paddle\fluid\dygraph\layers.py", line 915, in _dygraph_call_func outputs = self.forward(*inputs, kwargs) File "D:\project\GitHub\PaddleYOLO\PaddleYOLO\ppdet\modeling\heads\yolov5_head.py", line 132, in forward return self.loss(yolo_outputs, targets, self.anchors) File "D:\Program Files (x86)\python_package\PaddleYOLO\lib\site-packages\paddle\fluid\dygraph\layers.py", line 930, in call return self._dygraph_call_func(*inputs, *kwargs) File "D:\Program Files (x86)\python_package\PaddleYOLO\lib\site-packages\paddle\fluid\dygraph\layers.py", line 915, in _dygraph_call_func outputs = self.forward(inputs, kwargs) File "D:\project\GitHub\PaddleYOLO\PaddleYOLO\ppdet\modeling\losses\yolov5_loss.py", line 201, in forward tcls, tbox, indices, anch = self.build_targets_paddle(inputs, targets, anchors) File "D:\project\GitHub\PaddleYOLO\PaddleYOLO\ppdet\modeling\losses\yolov5_loss.py", line 282, in build_targets_paddle gxi = gain[[2, 3]] - gxy # inverse File "D:\Program Files (x86)\python_package\PaddleYOLO\lib\site-packages\paddle\fluid\dygraph\math_op_patch.py", line 303, in impl return math_op(self, other_var, 'axis', axis) RuntimeError: (PreconditionNotMet) The meta data must be valid when call the mutable data function. (at ..\paddle\phi\core\dense_tensor.cc:105) [operator < elementwise_sub > error]

复现环境 Environment

python 3.9.10 paddle 2.3.1 cuda 10.1 cuDNN 7.6

Bug描述确认 Bug description confirmation

• [X] 我确认已经提供了Bug复现步骤、代码改动说明、以及环境信息，确认问题是可以复现的。I confirm that the bug replication steps, code change instructions, and environment information have been provided, and the problem can be reproduced.

是否愿意提交PR？ Are you willing to submit a PR?

• [X] 我愿意提交PR！I'd like to help by submitting a PR!

[nemonameless]

win请安装更高版本paddle

[zhangxp12]

您好，paddle 升到 2.3.2 ，还是报这个错误

[nemonameless]

win的paddle需要更高版本，你可以直接安develop的paddle，打印版本号会显示是0.0.0