search

Tencent / ncnn

ncnn is a high-performance neural network inference framework optimized for the mobile platform
Other
20.11k stars 4.14k forks source link

CBAM module seems not working #3268

Open jeonginseo opened 2 years ago

jeonginseo commented 2 years ago

detail | 详细描述 | 詳細な説明

Hi, I implemented CBAM module with global average pooling and maxpooling. When I converted my onnx model to ncnn, it seemed be successful(sorry, I cannot upload the onnx model by private issue). But It gives me -nan values when I infer the model (I checked python and C++)

What can I do for this? Thanks

Attached file is converted .param file res34_CBAM.param.txt e

nihui commented 1 month ago

针对onnx模型转换的各种问题，推荐使用最新的pnnx工具转换到ncnn In view of various problems in onnx model conversion, it is recommended to use the latest pnnx tool to convert your model to ncnn

pip install pnnx
pnnx model.onnx inputshape=[1,3,224,224]

详细参考文档 Detailed reference documentation https://github.com/pnnx/pnnx https://github.com/Tencent/ncnn/wiki/use-ncnn-with-pytorch-or-onnx#how-to-use-pnnx