Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

yolov7导出onnx模型 #1

Open
Ellsionjeep opened this issue Nov 30, 2022 · 7 comments
Open

yolov7导出onnx模型 #1

Ellsionjeep opened this issue Nov 30, 2022 · 7 comments

Comments

@Ellsionjeep
Copy link

感谢作者提供的车牌检测的代码,请问作者能不能分享一下用yolv7实现的车牌检测转成onnx模型,我尝试着使用您在yolov5实现车牌检测中,将权重转成onnx,但在进行onnxruntime推理时,出现以下索引错误,发生异常: IndexError
index 4 is out of bounds for axis 1 with size 3

@we0091234
Copy link
Owner

感谢作者提供的车牌检测的代码,请问作者能不能分享一下用yolv7实现的车牌检测转成onnx模型,我尝试着使用您在yolov5实现车牌检测中,将权重转成onnx,但在进行onnxruntime推理时,出现以下索引错误,发生异常: IndexError index 4 is out of bounds for axis 1 with size 3
python models/export.py --weights weights/plate_detect.pt --grid

@Ellsionjeep
Copy link
Author

感谢作者提供的车牌检测的代码,请问作者能不能分享一下用yolv7实现的车牌检测转成onnx模型,我尝试着使用您在yolov5实现车牌检测中,将权重转成onnx,但在进行onnxruntime推理时,出现以下索引错误,发生异常: IndexError index 4 is out of bounds for axis 1 with size 3
python models/export.py --weights weights/plate_detect.pt --grid

好的,非常感谢作者的提供,大力支持

@Ellsionjeep
Copy link
Author

再次打扰作者一下,我用您给的权重导出onnx后,在yolo_pose_onnx_inference.py的post_process函数中出现The truth value of an array with more than one element is ambiguous. Use a.any() or a.all(),我打印了det_scores[idx].shape 输出是(19,)这一部分维度是否有问题

@we0091234
Copy link
Owner

再次打扰作者一下,我用您给的权重导出onnx后,在yolo_pose_onnx_inference.py的post_process函数中出现The truth value of an array with more than one element is ambiguous. Use a.any() or a.all(),我打印了det_scores[idx].shape 输出是(19,)这一部分维度是否有问题

onnx 推理的代码我还没跑过,19是没问题的,因为 xywh score这就是5个了,然后有两类2个分类得分,4个关键点,每个关键点包括x,y,score三个数,那么最后就是5+2+12=19

@Ellsionjeep
Copy link
Author

再次打扰作者一下,我用您给的权重导出onnx后,在yolo_pose_onnx_inference.py的post_process函数中出现The truth value of an array with more than one element is ambiguous. Use a.any() or a.all(),我打印了det_scores[idx].shape 输出是(19,)这一部分维度是否有问题

onnx 推理的代码我还没跑过,19是没问题的,因为 xywh score这就是5个了,然后有两类2个分类得分,4个关键点,每个关键点包括x,y,score三个数,那么最后就是5+2+12=19

我还打印了det_bboxes是(1,4,19)det_scores是(1,19),det_labels是(1,19),kpts是(1,25194,19),就感觉这个onnx推理跑不通

@we0091234
Copy link
Owner

再次打扰作者一下,我用您给的权重导出onnx后,在yolo_pose_onnx_inference.py的post_process函数中出现The truth value of an array with more than one element is ambiguous. Use a.any() or a.all(),我打印了det_scores[idx].shape 输出是(19,)这一部分维度是否有问题

onnx 推理的代码我还没跑过,19是没问题的,因为 xywh score这就是5个了,然后有两类2个分类得分,4个关键点,每个关键点包括x,y,score三个数,那么最后就是5+2+12=19

我还打印了det_bboxes是(1,4,19)det_scores是(1,19),det_labels是(1,19),kpts是(1,25194,19),就感觉这个onnx推理跑不通

找时间写个onnx推理的代码

@miracletrespasser
Copy link

export onnx模型时出现了这个错误 NYI:Named tensors are not supported with the tracer, 可能是在调用models/export.py 的第80行导致的,nn.Squential不支持命名向量,请问目前有什么好的解决办法吗?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants