Restructuring demos (#531)

Summary:
Pull Request resolved: #531

-----
## Confusion arising from the current code structure:

Reference to current code structure: https://github.com/pytorch/executorch/tree/main/examples

The current code structure under `executorch/examples` mixes different levels of detail. For instance, the directories `backend/`, `quantization/`, and `export/` are intended for component-level demo code, such as showcasing quantization/export/delegation workflows or providing sample backend/quantizer/partitioner implementations. Assembling them to create a AOT path should not belong to any of it. However, the current structure combines recipes and individual components:
  - All contents of `backend/` are exclusively tailored for [xnnpack](https://github.com/pytorch/executorch/tree/main/examples/backend).
  - [export+delegate+composite](https://github.com/pytorch/executorch/blob/main/examples/export/export_and_delegate.py) and [bundled program example](https://github.com/pytorch/executorch/blob/main/examples/export/export_bundled_program.py) are placed in `export/`
  - Different runtimes are placed haphazardly., with items like [executor_runner](https://github.com/pytorch/executorch/tree/main/examples/executor_runner), [bundled_executor_runner](https://github.com/pytorch/executorch/tree/main/examples/bundled_executor_runner), and [xnn_executor_runner](https://github.com/pytorch/executorch/tree/main/examples/backend)
  - Backend-specific demos are scattered in the top-level directory, such as [arm's demo](https://github.com/pytorch/executorch/tree/main/examples/arm) and [partitioner/quantizer demo](https://github.com/pytorch/executorch/tree/main/examples/example_quantizer_and_delegate)

Issues that are not directly tie to code restructuring but will need to be addressed as well. **Non-functional code/demo** (fix or hide):
  - [Unit tests](https://github.com/pytorch/executorch/tree/main/examples/export/test) under the `export/` directory are not runnable in the OSS environment.
  - The [example_quantizer_and_delegate](https://github.com/pytorch/executorch/tree/main/examples/example_quantizer_and_delegate) is not functioning and should be rectified or hidden.
  - The [arm's demo](https://github.com/pytorch/executorch/tree/main/examples/arm) is not runnable directly w/o additional guide.

-----

## Proposed code structure:
organize demos in a way that is easier for users to understand how pieces are put together to create an e2e experience. We can go with this "vertical" structure if we could come up and agree on guidelines of whether a new demo should be considered as "adding a new demo" or "extending an existing demo".

The purpose is to prevent randomly dumping demos in the top-level executorch/examples folder with boilerplate code. While this doesn't violate the rule of structuring demos vertically for user understanding, , but it literally has no structure at all, and it will quickly grow of out our maintenance.

In my opinion, it would make sense to consider the following cases as **"adding a new demo"**:
  1. Demo of a new 1st/3rd party backend, e.g. portable (default backend), xnnpack, arm, coreml, etc.
  2. Demo of toolings, e.g. selective build, productivity sdk, etc.
  3. Target specific apps, e.g. playground apps for Android/iOS/Embedded

The rests will be considered as **"extending an existing demo"**, for example:
  - Demo of composibitlity, e.g. delegate and composite, whole/partial graph lowering, custom ops
  - Demo of specific components, e.g. quantization workflow, 2-stage export flow, example quantizer/partitioner, etc.

After PTC we will spend more efforts on expanding and polishing the two major 1st-party demos (`examples/portable/` & `examples/xnnpack/`) we own in the long run. Maybe `examples/sdk` as well.

So with the guidelines, the new top-level structure will looks like:

```
executorch/examples/
├── README.md    # top-level description for demos and folder structures
├── models/
├── xnnpack/     # 1p e2e
├── portable/    # 1p e2e
├── arm/          # 3p backend e2e
├── qualcomm/     # 3p backend e2e
├── apple/        # 3p backend e2e
├── third-party/    # for the short-term, we can assume a centralized place for all third-party libs required by demos, e.g. sam, llama, etc.
├── selective_build/
├── ios_demo_apps/
├── android_demo_apps/   # future
├── productivity_sdk/    # future
└── utils/               # future
```

If we zoom in to `examples/portable/`, it will look like:
```
executorch/examples/portable/
├── executor_runner/
├── bundled_executor_runner/
├── custom_ops/
├── scripts/
│   ├── export_and_delegate.py
│   ├── export_bundled_program.py
│   └── export.py
├── test/
├── __init__.py
├── utils.py
└── README.md
```

If we zoom in to `examples/xnnpack/`, it will look like:
```
executorch/examples/xnnpack/
├── quantization/
│   ├── example.py
│   ├── TARGETS
│   ├── test_quantize.sh
│   └── utils.py
├── __init__.py
├── aot_compiler.py
└── README.md
```

Reviewed By: mergennachin

Differential Revision: D49714823

fbshipit-source-id: 592d94b1e0269b14641a6a98d2dc81be21e77734

Author: guangy10

.ci/scripts/gather_test_models.py

@@ -10,7 +10,7 @@
 from typing import Any
 
 from examples.models import MODEL_NAME_TO_MODEL
-from examples.recipes.xnnpack_optimization import MODEL_NAME_TO_OPTIONS
+from examples.xnnpack import MODEL_NAME_TO_OPTIONS
 
 # NB: Skip buck2 on MacOS to cut down the number of combinations we
 # need to run there as the number of MacOS runner is limited. Buck2
@@ -80,8 +80,7 @@ def export_models_for_ci() -> None:
         }
         delegation_configs = {
             False,
-            name in MODEL_NAME_TO_OPTIONS
-            and MODEL_NAME_TO_OPTIONS[name].xnnpack_delegation,
+            name in MODEL_NAME_TO_OPTIONS and MODEL_NAME_TO_OPTIONS[name].delegation,
         }
         for build_tool in BUILD_TOOLS.keys():
             if target_os not in BUILD_TOOLS[build_tool]:

.ci/scripts/test.sh

@@ -59,11 +59,11 @@ test_model() {
     cd ../../..
   fi
 
-  "${PYTHON_EXECUTABLE}" -m examples.export.export_example --model_name="${MODEL_NAME}"
+  "${PYTHON_EXECUTABLE}" -m examples.portable.scripts.export --model_name="${MODEL_NAME}"
 
   # Run test model
   if [[ "${BUILD_TOOL}" == "buck2" ]]; then
-    buck2 run //examples/executor_runner:executor_runner -- --model_path "./${MODEL_NAME}.pte"
+    buck2 run //examples/portable/executor_runner:executor_runner -- --model_path "./${MODEL_NAME}.pte"
   elif [[ "${BUILD_TOOL}" == "cmake" ]]; then
     if [[ ! -f ${CMAKE_OUTPUT_DIR}/executor_runner ]]; then
       build_cmake_executor_runner
@@ -98,24 +98,24 @@ test_model_with_xnnpack() {
 
   # Quantization-only
   if [[ ${WITH_QUANTIZATION} == true ]] && [[ ${WITH_DELEGATION} == false ]]; then
-    bash examples/quantization/test_quantize.sh "${BUILD_TOOL}" "${MODEL_NAME}"
+    bash examples/xnnpack/quantization/test_quantize.sh "${BUILD_TOOL}" "${MODEL_NAME}"
     exit 0
   fi
 
   # Delegation
   if [[ ${WITH_QUANTIZATION} == true ]]; then
     SUFFIX="q8"
-    "${PYTHON_EXECUTABLE}" -m examples.backend.xnnpack_examples --model_name="${MODEL_NAME}" --delegate --quantize
+    "${PYTHON_EXECUTABLE}" -m examples.xnnpack.aot_compiler --model_name="${MODEL_NAME}" --delegate --quantize
   else
     SUFFIX="fp32"
-    "${PYTHON_EXECUTABLE}" -m examples.backend.xnnpack_examples --model_name="${MODEL_NAME}" --delegate
+    "${PYTHON_EXECUTABLE}" -m examples.xnnpack.aot_compiler --model_name="${MODEL_NAME}" --delegate
   fi
 
   OUTPUT_MODEL_PATH="${MODEL_NAME}_xnnpack_${SUFFIX}.pte"
 
   # Run test model
   if [[ "${BUILD_TOOL}" == "buck2" ]]; then
-    buck2 run //examples/backend:xnn_executor_runner -- --model_path "${OUTPUT_MODEL_PATH}"
+    buck2 run //examples/xnnpack:xnn_executor_runner -- --model_path "${OUTPUT_MODEL_PATH}"
   elif [[ "${BUILD_TOOL}" == "cmake" ]]; then
     if [[ ! -f ${CMAKE_OUTPUT_DIR}/backends/xnnpack/xnn_executor_runner ]]; then
       build_cmake_xnn_executor_runner
@@ -129,15 +129,15 @@ test_model_with_xnnpack() {
 
 test_demo_backend_delegation() {
   echo "Testing demo backend delegation on AddMul"
-  "${PYTHON_EXECUTABLE}" -m examples.export.export_and_delegate  --option "composite"
-  "${PYTHON_EXECUTABLE}" -m examples.export.export_and_delegate  --option "partition"
-  "${PYTHON_EXECUTABLE}" -m examples.export.export_and_delegate  --option "whole"
+  "${PYTHON_EXECUTABLE}" -m examples.portable.scripts.export_and_delegate  --option "composite"
+  "${PYTHON_EXECUTABLE}" -m examples.portable.scripts.export_and_delegate  --option "partition"
+  "${PYTHON_EXECUTABLE}" -m examples.portable.scripts.export_and_delegate  --option "whole"
 
   # Run test model
   if [[ "${BUILD_TOOL}" == "buck2" ]]; then
-    buck2 run //examples/executor_runner:executor_runner -- --model_path "./composite_model.pte"
-    buck2 run //examples/executor_runner:executor_runner -- --model_path "./partition_lowered_model.pte"
-    buck2 run //examples/executor_runner:executor_runner -- --model_path "./whole.pte"
+    buck2 run //examples/portable/executor_runner:executor_runner -- --model_path "./composite_model.pte"
+    buck2 run //examples/portable/executor_runner:executor_runner -- --model_path "./partition_lowered_model.pte"
+    buck2 run //examples/portable/executor_runner:executor_runner -- --model_path "./whole.pte"
   elif [[ "${BUILD_TOOL}" == "cmake" ]]; then
     if [[ ! -f ${CMAKE_OUTPUT_DIR}/executor_runner ]]; then
       build_cmake_executor_runner

.ci/scripts/utils.sh

@@ -70,7 +70,7 @@ install_flatc_from_source() {
 
 build_executorch_runner_buck2() {
   # Build executorch runtime with retry as this step is flaky on macos CI
-  retry buck2 build //examples/executor_runner:executor_runner
+  retry buck2 build //examples/portable/executor_runner:executor_runner
 }
 
 build_executorch_runner_cmake() {

.github/workflows/pull.yml

@@ -58,7 +58,7 @@ jobs:
         MODEL_NAME=${{ matrix.model }}
         BUILD_TOOL=${{ matrix.build-tool }}
         XNNPACK_QUANTIZATION=${{ matrix.xnnpack_quantization }}
-        XNNPACK_DELEGATION=${{ matrix.xnnpack_delegation }}
+        XNNPACK_DELEGATION=${{ matrix.delegation }}
         DEMO_BACKEND_DELEGATION=${{ matrix.demo_backend_delegation }}
 
         PYTHON_EXECUTABLE=python bash .ci/scripts/setup-linux.sh "${BUILD_TOOL}"
@@ -87,7 +87,7 @@ jobs:
         BUILD_TOOL=${{ matrix.build-tool }}
         PYTHON_EXECUTABLE=python bash .ci/scripts/setup-linux.sh "${BUILD_TOOL}"
         # Test custom ops
-        PYTHON_EXECUTABLE=python bash examples/custom_ops/test_custom_ops.sh "${BUILD_TOOL}"
+        PYTHON_EXECUTABLE=python bash examples/portable/custom_ops/test_custom_ops.sh "${BUILD_TOOL}"
 
   test-selective-build-linux:
     name: test-selective-build-linux

.github/workflows/trunk.yml

@@ -55,7 +55,7 @@ jobs:
         MODEL_NAME=${{ matrix.model }}
         BUILD_TOOL=${{ matrix.build-tool }}
         XNNPACK_QUANTIZATION=${{ matrix.xnnpack_quantization }}
-        XNNPACK_DELEGATION=${{ matrix.xnnpack_delegation }}
+        XNNPACK_DELEGATION=${{ matrix.delegation }}
         DEMO_BACKEND_DELEGATION=${{ matrix.demo_backend_delegation }}
 
         # Setup MacOS dependencies as there is no Docker support on MacOS atm
@@ -85,7 +85,7 @@ jobs:
         PYTHON_EXECUTABLE=python bash .ci/scripts/setup-macos.sh "${BUILD_TOOL}"
 
         # Build and test custom ops
-        PYTHON_EXECUTABLE=python bash examples/custom_ops/test_custom_ops.sh "${BUILD_TOOL}"
+        PYTHON_EXECUTABLE=python bash examples/portable/custom_ops/test_custom_ops.sh "${BUILD_TOOL}"
         popd
 
   test-selective-build-macos:

CMakeLists.txt

@@ -93,8 +93,8 @@ set(CMAKE_CXX_FLAGS_DEBUG "-O0 -g")
 
 # Option to register custom operator `my_ops::mul3` or `my_ops::mul4` or no
 # custom ops at all. Custom ops are defined in
-# `examples/custom_ops/custom_ops_1.py` and
-# `examples/custom_ops/custom_ops_2.cpp`.
+# `examples/portable/custom_ops/custom_ops_1.py` and
+# `examples/portable/custom_ops/custom_ops_2.cpp`.
 option(
   REGISTER_EXAMPLE_CUSTOM_OP
   "Register whether custom op 1 (my_ops::mul3) or custom op 2 (my_ops::mul4) \
@@ -296,7 +296,7 @@ if(EXECUTORCH_BUILD_EXECUTOR_RUNNER)
 
   # Generate custom_ops_lib based on REGISTER_EXAMPLE_CUSTOM_OP
   if(REGISTER_EXAMPLE_CUSTOM_OP EQUAL 1 OR REGISTER_EXAMPLE_CUSTOM_OP EQUAL 2)
-    add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/examples/custom_ops)
+    add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/examples/portable/custom_ops)
     list(APPEND _executor_runner_libs custom_ops_lib)
   endif()
 

examples/example_quantizer_and_delegate/README.md renamed to backends/example/README.md

@@ -8,7 +8,7 @@ python_library(
     ],
     deps = [
         "//caffe2:torch",
-        "//executorch/examples/example_quantizer_and_delegate/example_operators:example_operators_lib",
+        "//executorch/backends/example/example_operators:example_operators_lib",
     ],
 )
 
@@ -18,7 +18,7 @@ python_library(
         "example_backend.py",
     ],
     deps = [
-        "//executorch/examples/example_quantizer_and_delegate/example_backend_delegate_passes:lib",
+        "//executorch/backends/example/example_backend_delegate_passes:lib",
         "//executorch/exir/backend:backend_details",
         "//executorch/exir/backend:compile_spec_schema",
     ],
@@ -32,7 +32,7 @@ python_library(
     deps = [
         ":example_backend",
         "//caffe2:torch",
-        "//executorch/examples/example_quantizer_and_delegate/example_operators:example_operators_lib",
+        "//executorch/backends/example/example_operators:example_operators_lib",
         "//executorch/exir:graph_module",
         "//executorch/exir/backend:partitioner",
         "//executorch/exir/backend/canonical_partitioners:canonical_partitioner_lib",

examples/example_quantizer_and_delegate/TARGETS renamed to backends/example/TARGETS

@@ -7,10 +7,10 @@
 import copy
 from typing import final, List
 
-from executorch.examples.example_quantizer_and_delegate.example_backend_delegate_passes.merge_to_dim_pass import (
+from executorch.backends.example.example_backend_delegate_passes.merge_to_dim_pass import (
     MergeToDimPass,
 )
-from executorch.examples.example_quantizer_and_delegate.example_backend_delegate_passes.permute_memory_formats_pass import (
+from executorch.backends.example.example_backend_delegate_passes.permute_memory_formats_pass import (
     PermuteMemoryFormatsPass,
 )
 

examples/example_quantizer_and_delegate/diagrams/delegate.png renamed to backends/example/diagrams/delegate.png

@@ -8,7 +8,7 @@ python_library(
     ],
     deps = [
         "//caffe2:torch",
-        "//executorch/examples/example_quantizer_and_delegate/example_operators:example_operators_lib",
+        "//executorch/backends/example/example_operators:example_operators_lib",
         "//executorch/exir:dim_order_utils",
         "//executorch/exir:pass_base",
         "//executorch/exir/dialects:lib",

examples/example_quantizer_and_delegate/diagrams/memory_permute.png renamed to backends/example/diagrams/memory_permute.png

@@ -7,9 +7,7 @@
 from itertools import chain
 
 import torch
-from executorch.examples.example_quantizer_and_delegate.example_operators.ops import (
-    module_to_annotator,
-)
+from executorch.backends.example.example_operators.ops import module_to_annotator
 from executorch.exir.dialects._ops import ops as exir_ops
 from executorch.exir.dim_order_utils import get_dim_order
 from executorch.exir.pass_base import ExportPass, PassResult

examples/example_quantizer_and_delegate/diagrams/quantize_delegate.png renamed to backends/example/diagrams/quantize_delegate.png

@@ -7,10 +7,8 @@
 from dataclasses import dataclass
 
 import torch
-from executorch.examples.example_quantizer_and_delegate.example_operators.op_base import (
-    OpBase,
-)
-from executorch.examples.example_quantizer_and_delegate.example_operators.utils import (
+from executorch.backends.example.example_operators.op_base import OpBase
+from executorch.backends.example.example_operators.utils import (
     _annotate_nodes,
     _nodes_are_annotated,
 )

examples/example_quantizer_and_delegate/example_backend.py renamed to backends/example/example_backend.py

@@ -7,10 +7,8 @@
 from dataclasses import dataclass
 
 import torch
-from executorch.examples.example_quantizer_and_delegate.example_operators.op_base import (
-    OpBase,
-)
-from executorch.examples.example_quantizer_and_delegate.example_operators.utils import (
+from executorch.backends.example.example_operators.op_base import OpBase
+from executorch.backends.example.example_operators.utils import (
     _annotate_nodes,
     _nodes_are_annotated,
 )

examples/example_quantizer_and_delegate/example_backend_delegate_passes/TARGETS renamed to backends/example/example_backend_delegate_passes/TARGETS

@@ -7,10 +7,8 @@
 from dataclasses import dataclass
 
 import torch
-from executorch.examples.example_quantizer_and_delegate.example_operators.op_base import (
-    OpBase,
-)
-from executorch.examples.example_quantizer_and_delegate.example_operators.utils import (
+from executorch.backends.example.example_operators.op_base import OpBase
+from executorch.backends.example.example_operators.utils import (
     _annotate_nodes,
     _nodes_are_annotated,
 )

examples/example_quantizer_and_delegate/example_backend_delegate_passes/merge_to_dim_pass.py renamed to backends/example/example_backend_delegate_passes/merge_to_dim_pass.py

@@ -7,10 +7,8 @@
 from dataclasses import dataclass
 
 import torch
-from executorch.examples.example_quantizer_and_delegate.example_operators.op_base import (
-    OpBase,
-)
-from executorch.examples.example_quantizer_and_delegate.example_operators.utils import (
+from executorch.backends.example.example_operators.op_base import OpBase
+from executorch.backends.example.example_operators.utils import (
     _annotate_nodes,
     _nodes_are_annotated,
 )

examples/example_quantizer_and_delegate/example_backend_delegate_passes/permute_memory_formats_pass.py renamed to backends/example/example_backend_delegate_passes/permute_memory_formats_pass.py

@@ -7,10 +7,8 @@
 from dataclasses import dataclass
 
 import torch
-from executorch.examples.example_quantizer_and_delegate.example_operators.op_base import (
-    OpBase,
-)
-from executorch.examples.example_quantizer_and_delegate.example_operators.utils import (
+from executorch.backends.example.example_operators.op_base import OpBase
+from executorch.backends.example.example_operators.utils import (
     _annotate_nodes,
     _nodes_are_annotated,
 )

examples/example_quantizer_and_delegate/example_operators/TARGETS renamed to backends/example/example_operators/TARGETS

@@ -7,10 +7,8 @@
 from dataclasses import dataclass
 
 import torch
-from executorch.examples.example_quantizer_and_delegate.example_operators.op_base import (
-    OpBase,
-)
-from executorch.examples.example_quantizer_and_delegate.example_operators.utils import (
+from executorch.backends.example.example_operators.op_base import OpBase
+from executorch.backends.example.example_operators.utils import (
     _annotate_nodes,
     _nodes_are_annotated,
 )

examples/example_quantizer_and_delegate/example_operators/adaptive_avg_pool2d.py renamed to backends/example/example_operators/adaptive_avg_pool2d.py

@@ -7,10 +7,8 @@
 from dataclasses import dataclass
 
 import torch
-from executorch.examples.example_quantizer_and_delegate.example_operators.op_base import (
-    OpBase,
-)
-from executorch.examples.example_quantizer_and_delegate.example_operators.utils import (
+from executorch.backends.example.example_operators.op_base import OpBase
+from executorch.backends.example.example_operators.utils import (
     _annotate_nodes,
     _nodes_are_annotated,
 )

examples/example_quantizer_and_delegate/example_operators/add.py renamed to backends/example/example_operators/add.py

@@ -6,28 +6,16 @@
 
 from collections import OrderedDict
 
-from executorch.examples.example_quantizer_and_delegate.example_operators.adaptive_avg_pool2d import (
+from executorch.backends.example.example_operators.adaptive_avg_pool2d import (
     AdaptiveAvgPool2dNode,
 )
 
-from executorch.examples.example_quantizer_and_delegate.example_operators.add import (
-    AddNode,
-)
-from executorch.examples.example_quantizer_and_delegate.example_operators.conv2d import (
-    Conv2DNode,
-)
-from executorch.examples.example_quantizer_and_delegate.example_operators.conv_relu import (
-    ConvReluNode,
-)
-from executorch.examples.example_quantizer_and_delegate.example_operators.dropout import (
-    DropOutNode,
-)
-from executorch.examples.example_quantizer_and_delegate.example_operators.flatten import (
-    FlattenNode,
-)
-from executorch.examples.example_quantizer_and_delegate.example_operators.linear import (
-    LinearNode,
-)
+from executorch.backends.example.example_operators.add import AddNode
+from executorch.backends.example.example_operators.conv2d import Conv2DNode
+from executorch.backends.example.example_operators.conv_relu import ConvReluNode
+from executorch.backends.example.example_operators.dropout import DropOutNode
+from executorch.backends.example.example_operators.flatten import FlattenNode
+from executorch.backends.example.example_operators.linear import LinearNode
 
 # The ordering of this is important as the quantizer will try to match the patterns in this order.
 # That's why we want to match the fused patterns first and then the non-fused ones.

examples/example_quantizer_and_delegate/example_operators/conv2d.py renamed to backends/example/example_operators/conv2d.py

@@ -7,12 +7,8 @@
 from typing import Dict, final
 
 import torch
-from executorch.examples.example_quantizer_and_delegate.example_backend import (
-    TosaBackend,
-)
-from executorch.examples.example_quantizer_and_delegate.example_operators.ops import (
-    module_to_annotator,
-)
+from executorch.backends.example.example_backend import TosaBackend
+from executorch.backends.example.example_operators.ops import module_to_annotator
 from executorch.exir.backend.canonical_partitioners.pattern_op_partitioner import (
     generate_partitions_from_list_of_nodes,
 )

examples/example_quantizer_and_delegate/example_operators/conv_relu.py renamed to backends/example/example_operators/conv_relu.py

@@ -8,9 +8,7 @@
 from typing import List
 
 import torch
-from executorch.examples.example_quantizer_and_delegate.example_operators.ops import (
-    module_to_annotator,
-)
+from executorch.backends.example.example_operators.ops import module_to_annotator
 from torch import fx
 from torch.ao.quantization.observer import HistogramObserver, MinMaxObserver
 from torch.ao.quantization.pt2e.graph_utils import find_sequential_partitions

examples/example_quantizer_and_delegate/example_operators/dropout.py renamed to backends/example/example_operators/dropout.py

@@ -10,12 +10,8 @@
 import torch
 import torch._export as export
 from executorch import exir
-from executorch.examples.example_quantizer_and_delegate.example_partitioner import (
-    ExamplePartitioner,
-)
-from executorch.examples.example_quantizer_and_delegate.example_quantizer import (
-    ExampleQuantizer,
-)
+from executorch.backends.example.example_partitioner import ExamplePartitioner
+from executorch.backends.example.example_quantizer import ExampleQuantizer
 from executorch.exir.backend.backend_api import to_backend
 
 from executorch.exir.backend.canonical_partitioners.duplicate_dequant_node_pass import (

examples/example_quantizer_and_delegate/example_operators/flatten.py renamed to backends/example/example_operators/flatten.py

@@ -13,7 +13,7 @@ excludes = [
 
 [targets.executor_runner]
 buck_targets = [
-  "//examples/executor_runner:executor_runner",
+  "//examples/portable/executor_runner:executor_runner",
 ]
 filters = [
   ".cpp$",
@@ -104,7 +104,7 @@ deps = [
 
 [targets.xnn_executor_runner]
 buck_targets = [
-  "//examples/backend:xnn_executor_runner",
+  "//examples/xnnpack:xnn_executor_runner",
 ]
 filters = [
   ".cpp$",