symbol lookup error: libtorch_cpu.so

[phys-chem]

When I run pair nequip with compiled LAMMPS, I have this error:
symbol lookup error: libtorch_cpu.so: undefined symbol: _ZNK3c1010TensorImpl22is_strides_like_customENS_12MemoryFormatE

However, I can run the LAMMPS/example/melt successfully. Do you know how to solve it? Thank you very much!

[phys-chem]

When I run pair nequip with compiled LAMMPS, I have this error: symbol lookup error: libtorch_cpu.so: undefined symbol: _ZNK3c1010TensorImpl22is_strides_like_customENS_12MemoryFormatE

However, I can run the LAMMPS/example/melt successfully. Do you know how to solve it? Thank you very much!

It is the problem of compiler. I have solved it. However, when I run LAMMPS, I still get an error:
terminate called after throwing an instance of 'std::out_of_range'
what(): Argument passed to at() was not in the map.

[phys-chem]

When I run pair nequip with compiled LAMMPS, I have this error: symbol lookup error: libtorch_cpu.so: undefined symbol: _ZNK3c1010TensorImpl22is_strides_like_customENS_12MemoryFormatE
However, I can run the LAMMPS/example/melt successfully. Do you know how to solve it? Thank you very much!

It is the problem of compiler. I have solved it. However, when I run LAMMPS, I still get an error: terminate called after throwing an instance of 'std::out_of_range' what(): Argument passed to at() was not in the map.

It seems that this error is caused by different branches of NequIP and LAMMPS pair. If I want to run NPT, should I used both stress branches for NequIP and pair-NequIP? Could you help me? Thank you very much!

[Linux-cpp-lisp]

That is correct that it's a branch mismatch. If you want to run with NPT, you need to have StressForceOutput on your model. (This can be done in main, see https://github.com/mir-group/nequip/blob/main/configs/minimal_stress.yaml, but you can also use nequip develop.)

[Linux-cpp-lisp]

Should be closed by #52.