core dumps when running examples

[mcthulhu]

I got core dumps when running both of your inference examples, e.g.

gptj_model_load: f16     = 1
gptj_model_load: ggml ctx size = 13334.86 MB
Illegal instruction (core dumped)

This was in a Linux Mint guest in VirtualBox with over 32 GB of memory. Any suggestions about what the problem might be?

[ggerganov]

Probably this can help:

https://forums.virtualbox.org/viewtopic.php?f=6&t=101917

What is the output of cat /proc/cpuinfo inside the VM?

[mcthulhu]

mike@mike-VirtualBox:~$ cat /proc/cpuinfo
processor : 0
vendor_id : GenuineIntel
cpu family : 6
model : 85
model name : Intel(R) Xeon(R) W-2135 CPU @ 3.70GHz
stepping : 4
microcode : 0xffffffff
cpu MHz : 3695.998
cache size : 8448 KB
physical id : 0
siblings : 3
core id : 0
cpu cores : 3
apicid : 0
initial apicid : 0
fpu : yes
fpu_exception : yes
cpuid level : 22
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx rdtscp lm constant_tsc rep_good nopl xtopology nonstop_tsc cpuid tsc_known_freq pni ssse3 cx16 pcid sse4_1 sse4_2 hypervisor lahf_lm invpcid_single pti fsgsbase invpcid md_clear flush_l1d arch_capabilities
bugs : cpu_meltdown spectre_v1 spectre_v2 spec_store_bypass l1tf mds swapgs itlb_multihit mmio_stale_data retbleed
bogomips : 7391.99
clflush size : 64
cache_alignment : 64
address sizes : 46 bits physical, 48 bits virtual
power management:

processor : 1
vendor_id : GenuineIntel
cpu family : 6
model : 85
model name : Intel(R) Xeon(R) W-2135 CPU @ 3.70GHz
stepping : 4
microcode : 0xffffffff
cpu MHz : 3695.998
cache size : 8448 KB
physical id : 0
siblings : 3
core id : 1
cpu cores : 3
apicid : 1
initial apicid : 1
fpu : yes
fpu_exception : yes
cpuid level : 22
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx rdtscp lm constant_tsc rep_good nopl xtopology nonstop_tsc cpuid tsc_known_freq pni ssse3 cx16 pcid sse4_1 sse4_2 hypervisor lahf_lm invpcid_single pti fsgsbase invpcid md_clear flush_l1d arch_capabilities
bugs : cpu_meltdown spectre_v1 spectre_v2 spec_store_bypass l1tf mds swapgs itlb_multihit mmio_stale_data retbleed
bogomips : 7391.99
clflush size : 64
cache_alignment : 64
address sizes : 46 bits physical, 48 bits virtual
power management:

processor : 2
vendor_id : GenuineIntel
cpu family : 6
model : 85
model name : Intel(R) Xeon(R) W-2135 CPU @ 3.70GHz
stepping : 4
microcode : 0xffffffff
cpu MHz : 3695.998
cache size : 8448 KB
physical id : 0
siblings : 3
core id : 2
cpu cores : 3
apicid : 2
initial apicid : 2
fpu : yes
fpu_exception : yes
cpuid level : 22
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx rdtscp lm constant_tsc rep_good nopl xtopology nonstop_tsc cpuid tsc_known_freq pni ssse3 cx16 pcid sse4_1 sse4_2 hypervisor lahf_lm invpcid_single pti fsgsbase invpcid md_clear flush_l1d arch_capabilities
bugs : cpu_meltdown spectre_v1 spectre_v2 spec_store_bypass l1tf mds swapgs itlb_multihit mmio_stale_data retbleed
bogomips : 7391.99
clflush size : 64
cache_alignment : 64
address sizes : 46 bits physical, 48 bits virtual
power management:

[ggerganov]

Yes, it is missing the avx and avx2 CPU flags. These are required when running ggml on x86 architectures.

[mcthulhu]

Thank you for the hints; I managed to disable Hyper-V and get the examples to work. /proc/cpuinfo now shows avx and avx2.