Segmentation fault (core dumped) while trainning

I tried to create at small example that triggers it but I wasn’t successful. Now I can’t reproduce it even in my original program, even though it happened consistently before… I will show the code here if I can trigger it again.

Let me know if you see it again :slight_smile:

Hi
I have got the same error:

Thread 73 "python" received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x7ffe0dcc4700 (LWP 19486)]
0x00007fffa22ed8d5 in ?? () from /usr/lib/x86_64-linux-gnu/libcuda.so.1
(gdb) bt
#0  0x00007fffa22ed8d5 in ?? () from /usr/lib/x86_64-linux-gnu/libcuda.so.1
#1  0x00007fffa243d914 in ?? () from /usr/lib/x86_64-linux-gnu/libcuda.so.1
#2  0x00007fffa23d9e80 in ?? () from /usr/lib/x86_64-linux-gnu/libcuda.so.1
#3  0x00007ffff7bc16ba in start_thread (arg=0x7ffe0dcc4700) at pthread_create.c:333
#4  0x00007ffff78f741d in clone () at ../sysdeps/unix/sysv/linux/x86_64/clone.S:109

could you please help me why does this problem occur?