Xuan Guo
Xuan Guo
What's the status of this issue? Is there any plan to support python 3.9?
I have same issue, WSL Ubuntu 20.04 with RTX 3090, neither tensorrt 8.4 or 8.2 work
> @hsuanguo Please, fill all details from the bug form. You skipped your Conan version, which is really important. Also, share your executed commands and full log if possible. Only...
Also, OpenCV failed using zlib as dependencies with every cross building cases, with the following error: ``` -- Conan: Using autogenerated FindZLIB.cmake -- Could NOT find ZLIB: Found unsuitable version...
Hi, thanks for the testing, I upgraded to conan 1.51.1, the old issue is gone, and I do have to pass the ndk_path like you did now, however, it still...
> @hsuanguo The test package is not prepared for your case, try the folder test_v1_package instead: > > ``` > conan create . 1.2.12@_/_ --profile:build=default --profile:host=../../profiles/android-armv8-clang9-api29_release.txt -c tools.android:ndk_path=/home/xuang/.conan/data/android-ndk/r21d/_/_/package/4db1be536558d833e52e862fd84d64d75c2b3656/ -tf test_v1_package...
I guess https://github.com/apache/tvm/pull/13269 does not fix the python 3.9 support completely?
应该是w,h https://github.com/PaddlePaddle/PaddleClas/blob/ef796960c0e4227db606d7683254a0d6352ef3f5/ppcls/data/preprocess/ops/operators.py#L314 整个repo几乎都是w,h
I upgraded to tensorRT 8.5.3, and the previous issue is gone now.
OK, 看代码的话貌似是没有优先级或者自动排除的机制, 那么就确实是两个优化器有作用于相同的参数的情况