search

cortex-lab / KiloSort

GPU code for spike sorting
GNU General Public License v2.0
175 stars 100 forks source link

Error running Kilosort! #225

Open parviz-gh opened 4 years ago

parviz-gh commented 4 years ago

hi, i have 2 sessions recording with Neuropixel . i did spike sorting using Kilosort2 separately , all was good. but when i concatenated them using SpikeGLX then run Kilosort2 on concatenated stream the following error was appeared : 'Error running Kilosort! an unexpected error occurred trying to lunch a kernel. CUDA error was: invalid configuration argument'

JannaAarse commented 3 years ago

Hey,

I have the same problem, but not only with concatenated files also with datasets that are recorded continuously (some of them are about 3 hours long) and when I try to run kilosort2.5 with them it takes forever (yesterday the preproccesing step took over 5 hours) and it always crashes when it is about to shift or optimizing the batches. The message I get is the following:

Error using gpuArray/nan An unexpected error occurred trying to launch a kernel. The CUDA error was: invalid configuration argument

Error in median(line 71) y=nan(s,'like',x)

Error in learnTemplates (line 248) toc,ibatch,niter,Nfilt,sum(nsp),median(mu),numel(st0),ndrop)

Error in learnAndSolve8b (line 35) rez= learnTemplates(rez,rez.iorig);

Error in main_kilosort25 (line 46) rez= learnAndSolve8b(rez,iseed);

How can I fix this problem ? Do you have any idea why this is happening ?

Best, Janna

mahdiramadan commented 2 years ago

Hi all! Any update on this issue and how to fix it?

JannaAarse commented 2 years ago

Hey, I did not really solve it , I remember I had problems with some recordings, there were flat lines (saturations) in the recording , those ones gave me struggles and I wasn't able to use them (but since the signal was saturated because of noise it is anyway questionable if I should use those recordings). I did throw them out and are not using recordings where I had problems with the grounding any more. I am sorry that I can't give a more satisfying answer.

Best, Janna