Update base for Update on "Dont quantize the current token for attention"

Differential Revision: [D63497872](https://our.internmc.facebook.com/intern/diff/D63497872/)

[ghstack-poisoned]

Author: kimishpatel

.ci/scripts/test_llama.sh

@@ -188,7 +188,7 @@ EXPORTED_MODEL_NAME="${EXPORTED_MODEL_NAME}.pte"
 echo "Exporting ${EXPORTED_MODEL_NAME}"
 EXPORT_ARGS="-c ${CHECKPOINT_FILE_NAME} -p ${PARAMS} -d ${DTYPE} -n ${EXPORTED_MODEL_NAME} -kv"
 if [[ "${XNNPACK}" == "ON" ]]; then
-  EXPORT_ARGS="${EXPORT_ARGS} -X -qmode 8da4w -G 128"
+  EXPORT_ARGS="${EXPORT_ARGS} -X --xnnpack-extended-ops -qmode 8da4w -G 128"
 fi
 if [[ "${CUSTOM}" == "ON" ]]; then
   EXPORT_ARGS="${EXPORT_ARGS} --use_sdpa_with_kv_cache"

backends/cadence/aot/compiler.py

@@ -30,7 +30,6 @@
 )
 from executorch.backends.transforms.remove_clone_ops import RemoveCloneOpsTransform
 from executorch.exir import EdgeCompileConfig, EdgeProgramManager, to_edge
-from torch._export import capture_pre_autograd_graph
 from torch.ao.quantization.pt2e.export_utils import model_is_exported
 from torch.ao.quantization.quantize_pt2e import convert_pt2e, prepare_pt2e
 
@@ -58,7 +57,7 @@ def convert_pt2(
     """
 
     # Export with dynamo
-    model_gm = capture_pre_autograd_graph(model, inputs)
+    model_gm = torch.export.export_for_training(model, inputs).module()
 
     if model_gm_has_SDPA(model_gm):  # pyre-fixme[6]
         # Decompose SDPA

backends/cadence/cadence_runner/CMakeLists.txt

@@ -44,7 +44,9 @@ find_package(
   gflags REQUIRED PATHS ${CMAKE_CURRENT_BINARY_DIR}/../../../third-party
 )
 
-add_executable(cadence_runner cadence_runner.cpp)
+add_executable(cadence_runner
+    ${EXECUTORCH_ROOT}/examples/devtools/example_runner/example_runner.cpp
+)
 target_compile_options(executorch INTERFACE -DET_EVENT_TRACER_ENABLED)
 
 add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/${TARGET_DIR}/operators)

backends/cadence/cadence_runner/TARGETS

@@ -9,8 +9,6 @@
 #include <executorch/backends/cadence/reference/kernels/kernels.h>
 
 #include <executorch/runtime/kernel/kernel_includes.h>
-#include <algorithm>
-#include <cmath>
 
 namespace impl {
 namespace reference {

backends/cadence/cadence_runner/cadence_runner.cpp

@@ -9,9 +9,7 @@
 #include <executorch/backends/cadence/reference/kernels/kernels.h>
 #include <executorch/runtime/kernel/kernel_includes.h>
 
-#include <algorithm>
 #include <cmath>
-#include <tuple>
 
 using Tensor = exec_aten::Tensor;
 using executorch::runtime::KernelRuntimeContext;