RuntimeError: cuda runtime error (30) : unknown error at ..\aten\src\THC\THCGeneral.cpp:87 while moving model to cuda

When i run “torch.cuda.current_device()” or move a model to cuda, i receive a cuda runtime error(30)
This is the traceback

nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2019 NVIDIA Corporation
Built on Fri_Feb__8_19:08:26_Pacific_Standard_Time_2019
Cuda compilation tools, release 10.1, V10.1.105

-----conda environment installations----
cudatoolkit || 10.0.130 || 0
cudnn || 7.3.1 || cuda10.0_0
pytorch || 1.0.1 || py3.7_cuda100_cudnn7_1 pytorch
torchvision || 0.2.2 || py_3 pytorch

i have two cuda installations, one in “C:\ program files” and also in “conda virtual environment”. This problem occurs while working in the conda environment. I have all drivers and installations up to date… but i can’t narrow down where the problem is. I even tried other solution available in the previous discussions.

These are my current environment path variables