Yi Zhang

@janeyx99

> Please keep the pinning of the specific NDK version. You can remove the installation if you want. It's hard to detect if NDK version changes introduce issues otherwise. It...

@thiagocrepaldi @liqunfu https://github.com/onnx/onnx/pull/4769 has been merged, but test_col2im_pads_cpu is still failed. Is it expected? ``` 2023-02-15T03:34:57.7065257Z ====================================================================== 2023-02-15T03:34:57.7065861Z FAIL: test_col2im_pads_cpu (__main__.OnnxBackendNodeModelTest) 2023-02-15T03:34:57.7066498Z ---------------------------------------------------------------------- 2023-02-15T03:34:57.7066827Z Traceback (most recent call last): 2023-02-15T03:34:57.7067409Z...

> @mszhanyi this PR removes some functionality we rely on, that ensures we get the correct version. Is that functionality difficult to support on Windows? What are the issues you...

Just a reminder, please don't forget to add --pure_pytorch false when you verify it with nbdev on Windows. ` fastrelease_conda_package --pure_python false`.

It's a mistake, I remember I took it as python version at some point. It's recovered now. I don't try to upload it. I think I have no permission and...

@jph00 is there more comments?

pip install onnxruntime==1.17.1 works on my machine. If you want to install onnxruntime-gpu (CUDA 12.x), please run `pip install onnxruntime-gpu --extra-index-url https://aiinfra.pkgs.visualstudio.com/PublicPackages/_packaging/onnxruntime-cuda-12/pypi/simple/` https://onnxruntime.ai/docs/install/

Did you run `" ...\...\vcvarsall.bat" x64 -vcvars_ver=14.31`

@dfalbel could you use randtemp.exe as the snippet https://github.com/pytorch/builder/blob/d0bc74cc363a9da5a8b6a40e883d40d25d050036/windows/build_pytorch.bat#L98-L114