I don't know how to do it.

[ytgcljj]

gtx-16@gtx16-Lenovo:~/Detectron-Cascade-RCNN$ sudo nvidia-docker run --rm -it detectron:c2-cuda9-cudnn7 python2 detectron/tests/test_batch_permutation_op.py No handlers could be found for logger "caffe2.python.net_drawer" net_drawer will not run correctly. Please install the correct dependencies. E0929 07:30:17.760017 1 init_intrinsics_check.cc:54] CPU feature avx is present on your machine, but the Caffe2 binary is not compiled with it. It means you may not get the full speed of your CPU. E0929 07:30:17.760035 1 init_intrinsics_check.cc:54] CPU feature avx2 is present on your machine, but the Caffe2 binary is not compiled with it. It means you may not get the full speed of your CPU. E0929 07:30:17.760040 1 init_intrinsics_check.cc:54] CPU feature fma is present on your machine, but the Caffe2 binary is not compiled with it. It means you may not get the full speed of your CPU. Found Detectron ops lib: /usr/local/caffe2_build/lib/libcaffe2_detectron_ops_gpu.so ..

Ran 2 tests in 0.726s

OK

[ytgcljj]

I closed it after I finished the command. How can I see the file? The difference between tools/infer_simple.py and test_batch_permutation_op.py

[kingwpf]

may be you don't need install caffe2 by compiling the source code, just install pytorch, the caffe2 will be installed automatically.