[ExecuTorch][llm] Add source-transform and Meta-kernel tests for MoE op (pytorch#21536)

This PR was created by the merge bot to help merge the original PR into
the main branch.
ghstack PR number: pytorch#21122 by
@digantdesai
^ Please use this as the source of truth for the PR details, comments,
and reviews
ghstack PR base:
https://github.com/pytorch/executorch/tree/gh/digantdesai/72/base
ghstack PR head:
https://github.com/pytorch/executorch/tree/gh/digantdesai/72/head
Merge bot PR base:
https://github.com/pytorch/executorch/tree/gh/digantdesai/71/orig
Merge bot PR head:
https://github.com/pytorch/executorch/tree/gh/digantdesai/72/orig
Differential Revision:
[D102381997](https://our.internmc.facebook.com/intern/diff/D102381997/)
@diff-train-skip-merge

Co-authored-by: Digant Desai <digantdesai@meta.com>

Author: digantdesai

.ci/scripts/test_llama_torchao_lowbit.sh

@@ -50,6 +50,9 @@ cmake -DPYTHON_EXECUTABLE=python \
     examples/models/llama
 cmake --build cmake-out/examples/models/llama -j16 --config Release
 
+${PYTHON_EXECUTABLE} -m pytest \
+    extension/llm/custom_ops/test_quantized_moe.py
+
 # Download stories llama110m artifacts
 download_stories_model_artifacts
 

extension/llm/custom_ops/BUCK

@@ -129,3 +129,22 @@ fbcode_target(_kind = runtime.python_binary,
     ],
     deps = [":bench_channelwise_gated_delta_rule_lib"],
 )
+
+fbcode_target(_kind = runtime.python_test,
+    name = "test_quantized_moe",
+    srcs = [
+        "test_quantized_moe.py",
+    ],
+    preload_deps = [
+        ":custom_ops_aot_lib_mkl_noomp",
+        ":custom_ops_aot_py",
+        "//pytorch/ao/torchao/csrc/cpu/shared_kernels/linear_8bit_act_xbit_weight:op_linear_8bit_act_xbit_weight_aten",
+    ],
+    deps = [
+        "//caffe2:torch",
+        "//executorch/extension/pybindings:portable_lib",
+        "//executorch/examples/models/llama:llama_transformer",
+        "//executorch/examples/models/llama:transformer_modules",
+        "//executorch/examples/models/llama:source_transformation",
+    ],
+)