Remove libOMP linking for experimental kernels

[Jack-Khuu]

Remake of #1822

---

As mentioned in pytorch/torchchat#1493, we no longer need to explicitly link to OMP libs.

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/1828

• 📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

❌ 2 New Failures

As of commit 4da1fed with merge base ffb4350 ():

NEW FAILURES - The following jobs have failed:

• Run TorchAO Experimental Tests / test-cpu-ops (macos-14) (gh)
  Process completed with exit code 1.
• Run TorchAO Experimental Tests / test-mps-ops (macos-m1-stable) (gh)
  Process completed with exit code 1.

This comment was automatically generated by Dr. CI and updates every 15 minutes.

[Jack-Khuu]

Looks like just removing the omp link isn't enough. It can't find some omp related ops

Undefined symbols for architecture arm64:
        "___kmpc_fork_call", referenced from:
             ...
        "_omp_get_num_threads", referenced from:
            _.omp_outlined. in linear_8bit_act_xbit_weight.cpp.o
            _.omp_outlined..19 in linear_8bit_act_xbit_weight.cpp.o
            _.omp_outlined..20 in linear_8bit_act_xbit_weight.cpp.o
            _.omp_outlined..21 in linear_8bit_act_xbit_weight.cpp.o
            _.omp_outlined. in op_embedding_xbit_aten.cpp.o
            _.omp_outlined..189 in op_embedding_xbit_aten.cpp.o
            _.omp_outlined..192 in op_embedding_xbit_aten.cpp.o
            ...
        "_omp_get_thread_num", referenced from:

Note: This is unrelated (most likely) to the failures on main

[metascroy]

I don't think this will work unless libomp.dylib is included in TORCH_LIBRARIES. CI is failing as well.