mmchen88

for data, label in data_loader: with torch.no_grad(): output = model(data) output = model(data).data.cpu().numpy()

1>------ 已启动生成: 项目: libcaffe, 配置: Release x64 ------ 1> ProtoCompile.cmd : Create proto temp directory "E:\caffe-windows-ms\windows\..\src\caffe\proto\temp" 1> ProtoCompile.cmd : Generating "E:\caffe-windows-ms\windows\..\src\caffe\proto\temp\caffe.pb.h" and "E:\caffe-windows-ms\windows\..\src\caffe\proto\temp\caffe.pb.cc" 1> ProtoCompile.cmd : Create proto include directory...

This is my construction of Classifier: Classifier::Classifier(const std::string& model_file, const std::string& trained_file) { #ifdef CPU_ONLY Caffe::set_mode(Caffe::CPU); #else Caffe::set_mode(Caffe::GPU); #endif /* Load the network. */ net_.reset(new Net(model_file, TEST)); net_->CopyTrainedLayersFrom(trained_file); CHECK_EQ(net_->num_inputs(), 1)...

the net can't be hybridized

2

when opt.hybridize=True AssertionError: Argument data must have NDArray type, but got Symbol resnet0_hybridsequential2_activation7 in the line: parts_2 = nd.split(x2, axis=2, num_outputs=2)

If the gpu id was 0, the result would be all right. If the gpu id was 1, or 2, or 3, the output of roi_align would be all zero....

数据增强中上下翻转角度不变对么？

1

我的感觉是上下翻转和左右翻转角度都应该变成补角呢？请赐教

请问x64设备上怎么运行v3版本的mnn的模型？

1

请问x64设备上怎么运行v3版本的mnn的模型？

onnx-runtime export error

1

Adapointr模型转换成onnx后，用onnx-runtime推理出来的结果是错的，请问怎么解决呢？