evangeb / yolov3 Goto Github PK

Hello, what is your configuration environment? Is it a model trained by calling yolov3 (Python) with libtorch?

when i run "./darknet data/ronaldo.jpg" always wrong!

Upsample Forward Layer
terminate called after throwing an instance of 'c10::Error'
what(): It is expected output_size equals to 2, but got size 1 (upsample_bilinear2d_out_cpu_template at /pytorch/aten/src/ATen/native/UpSampleBilinear2d.cpp:165)
frame #0: std::function<std::string ()>::operator()() const + 0x11 (0x7f8eaea5e441 in /home/tb/Downloads/libtorch/lib/libc10.so)
frame #1: c10::Error::Error(c10::SourceLocation, std::string const&) + 0x2a (0x7f8eaea5dd7a in /home/tb/Downloads/libtorch/lib/libc10.so)
frame #2: + 0xa760b4 (0x7f8ea4a610b4 in /home/tb/Downloads/libtorch/lib/libcaffe2.so)

What caused it?thank you very much!

你好，请问你配置的环境是什么？是用libTorch 调用yolov3 pytorch 训练的模型吗？

ubuntu18.04
pytorch 1.5.1
libtorch 1.2.0
cuda 10.0
cudnn 7.6.5

hello, there was an error occured when i compile the project. the error is:
error: ‘class torch::nn::BatchNormImpl’ has no member named ‘running_var’
batch_norm->running_var = batchNormVarTensor.view_as(batch_norm->running_var);