Installation Error for CUDA 10.1, PyTorch 1.3.0 and gcc version 9.3.0 in conda environment

I am trying to install NVIDIA apex in the conda environment with gcc version 9.3.0, Python 3.6, PyTorch 1.3.0, CUDA 10.1, nvidia 3060
I have cloned the git repository followed by the command:
pip install -v --disable-pip-version-check --no-cache-dir --global-option="--cpp_ext" --global-option="--cuda_ext" ./
However, some errors happened:
Unknown CUDA arch (8.6) or GPU not supported
Running setup.py install for apex ... error
ERROR: Command errored out with exit status 1: /home/pengxing/anaconda3/envs/nv/bin/python -u -c 'import sys, setuptools, tokenize; sys.argv[0] = '"'"'/tmp/pip-req-build-2d46c1yh/setup.py'"'"'; file='"'"'/tmp/pip-req-build-2d46c1yh/setup.py'"'"';f=getattr(tokenize, '"'"'open'"'"', open)(file);code=f.read().replace('"'"'\r\n'"'"', '"'"'\n'"'"');f.close();exec(compile(code, file, '"'"'exec'"'"'))' --cpp_ext --cuda_ext install --record /tmp/pip-record-kn5qcyng/install-record.txt --single-version-externally-managed --compile --install-headers /home/pengxing/anaconda3/envs/nv/include/python3.6m/apex Check the logs for full command output.
please help me,thanks

Start with the reported pip install command and the setup.py output, especially the “Unknown CUDA arch (8.6)” error. Check how this project handles GPU architectures for the stated PyTorch and CUDA versions, then verify whether a supported installation path can be demonstrated for this environment.