inplace_abn
inplace_abn copied to clipboard
In-Place Activated BatchNorm for Memory-Optimized Training of DNNs
When attempting to import Inplace-ABN version 1.1.0, I get the following error: ``` >>> from inplace_abn import InPlaceABN Traceback (most recent call last): File "", line 1, in File "/home/curttigges/miniconda3/envs/pytorch-dl/lib/python3.7/site-packages/inplace_abn/__init__.py",...
LookupError: setuptools-scm was unable to detect version for /home/pyl/pythonproject_pyl/inplace_abn_main. Make sure you're either building from a fully intact git repository or PyPI tarballs. Most other sources (such as GitHub's tarballs,...
If you want to install inplace_ABN based on this platform, execute the following command: pip install git+https://github.com/mapillary/inplace_abn.git#egg=inplace_abn     
Hi I'm trying to install setup.py and I'm getting: "error: command '/usr/local/cuda/bin/nvcc' failed with exit status 1" I have a valid path :/usr/local/cuda-10.1/bin/nvcc, but it always picks the path shown...
With environment torch==1.8.1 CUDA 10.1 failed the installation. After update CUDA to 11.1 successfully installed. So for those who got problems with installation especially nvcc error, maybe update the cuda...
When I do: from inplace_abn import InPlaceABN, InPlaceABNSync there is an error: /home/xx/.local/lib/python3.6/site-packages/inplace_abn-1.0.7-py3.6-linux-x86_64.egg/inplace_abn/_backend.cpython-36m-x86_64-linux-gnu.so: undefined symbol: _ZN3c105ErrorC1ENS_14SourceLocationERKSs Can anyone help me? Environment: cuda 10.0 pytorch 1.2 torchvision 0.4.0 gcc 7.4.0
hello, i got an import error: When `import inplace_abn` It tells : Traceback (most recent call last): File "", line 1, in File "/home/lib/python3.7/site-packages/inplace_abn/__init__.py", line 1, in from .abn import...
I am using Python 3.7.5, NVCC=10.1.243, GCC=8.4.0, pytorch = 1.7.1. When I try doing : pip install inplace-abn. Got this NVCC error. I check in /usr/bin/nvcc and the nvcc file...
Are there any plans to support Cuda 11.0+ versions?