Knet init error

[fatihtalu]

I installed Julia.
I installed CUDA, IJulia, Knet and many other libraries.
I ran Knet/example/LeNet.
The training was completed in 3 seconds using the GPU and the accuracy was 97%.
Everything was perfect.

However, half an hour ago, the Nvidia GPU driver update came, and after installing the update, when I open Julia and give the command "using Knet", it gives "Error: Init Error" and automatically closes Julia.

When I run the command pkg.test (" Knet ") I get an error similar to the following.

_start at. \ client.jl: 506 jfptr__start_58799.clone_1 at C: \ Users \ Administrator \ AppData \ Local \ Programs \ Julia 1.5.3 \ lib \ julia \ sys.dll (unknown line) jl_apply at / cygdrive / c / buildbot / worker / package_win64 / build / src \ julia.h: 1690 [inlined] true_main at /cygdrive/c/buildbot/worker/package_win64/build/ui\repl.c:106 wmain at /cygdrive/c/buildbot/worker/package_win64/build/ui\repl.c:227 __tmainCRTStartup at /usr/src/debug/mingw64-x86_64-runtime-8.0.0-1/crt\crtexe.c:328 mainCRTStartup at /usr/src/debug/mingw64-x86_64-runtime-8.0.0-1/crt\crtexe.c:212 BaseThreadInitThunk at C: \ Windows \ System32 \ KERNEL32.DLL (unknown line) RtlUserThreadStart at C: \ Windows \ SYSTEM32 \ ntdll.dll (unknown line) Allocations: 38888689 (Pool: 38875398; Big: 13291); GC: 33 ERROR: Package Knet errored during testing Stacktrace: [1] pkgerror (:: String) at C: \ buildbot \ worker \ package_win64 \ build \ usr \ share \ julia \ stdlib \ v1.5 \ Pkg \ src \ Types.jl: 52 [2] test (:: Pkg.Types.Context, :: Array {Pkg.Types.PackageSpec, 1); coverage :: Bool, julia_args :: Cmd, test_args :: Cmd, test_fn :: Nothing) at C: \ buildbot \ worker \ package_win64 \ build \ usr \ share \ julia \ stdlib \ v1.5 \ Pkg \ src \ Operations.jl: 1578

I removed Julia and reinstalled it didn't get any better.
I am using Windows 10, CUDA 11.1 and Nvidia RTX 2080 Ti

[lanxaustralis]

Replicated the same problem on Windows 10, CUDA 11.2 and Nvidia GTX 960M laptop. As mentioned earlier in this issue, the error may be due to an unsupported version of NVML. Yet, quite interestingly, if REPL still works after the error, running "using Knet" command successfully loads Knet. Even if this is not applicable for notebook files (as this process ends up with a dead kernel), you may try calling Knet again, until the issue is resolved.