Arm backend: Add and use ArmQuantizer

backends/arm/README.md

@@ -20,6 +20,10 @@ kend Architecture](#arm-backend-architecture). For examples of use see `executor
 - `tosa_mapping.py` - utilities for mapping edge dialect to TOSA
 - `tosa_quant_utils.py` - utilities for mapping quantization information to TOSA encoding
 
+Quantization:
+- `arm_quantizer.py` - Quantizer for Arm backend
+- `arm_quantizer_utils.py` - Utilities for quantization
+
 Runtime:
 - `runtime/ArmBackendEthosU.cpp` - The Arm backend implementation of the ExecuTorch runtime backend (PyTorchBackendInterface) for Ethos-U
 

backends/arm/arm_quantizer.py

@@ -1,9 +1,14 @@
 # Copyright (c) Meta Platforms, Inc. and affiliates.
+# Copyright 2024 Arm Limited and/or its affiliates.
 # All rights reserved.
 #
 # This source code is licensed under the BSD-style license found in the
 # LICENSE file in the root directory of this source tree.
 
+#
+# Quantizer for Arm backend
+#
+
 from __future__ import annotations
 
 import copy
@@ -14,6 +19,15 @@
 import torch
 import torch._dynamo as torchdynamo
 import torch.nn.functional as F
+
+from executorch.backends.arm.arm_quantizer_utils import (
+    _convert_scalars_to_attrs,
+    OP_TO_ANNOTATOR,
+    OperatorConfig,
+    OperatorPatternType,
+    propagate_annotation,
+    QuantizationConfig,
+)
 from torch.ao.quantization.fake_quantize import (
     FakeQuantize,
     FusedMovingAvgObsFakeQuantize,
@@ -31,20 +45,11 @@
 
 from torch.ao.quantization.quantizer import QuantizationSpec, Quantizer
 
-from torch.ao.quantization.quantizer.xnnpack_quantizer_utils import (
-    _convert_scalars_to_attrs,
-    OP_TO_ANNOTATOR,
-    OperatorConfig,
-    OperatorPatternType,
-    propagate_annotation,
-    QuantizationConfig,
-)
-
 from torch.fx import Node
 
 
 __all__ = [
-    "XNNPACKQuantizer",
+    "ArmQuantizer",
     "get_symmetric_quantization_config",
 ]
 
@@ -260,7 +265,7 @@ def not_module_type_or_name_filter(n: Node) -> bool:
     return not_module_type_or_name_filter
 
 
-class XNNPACKQuantizer(Quantizer):
+class ArmQuantizer(Quantizer):
     supported_config_and_operators = _get_supported_config_and_operators()
     STATIC_QAT_ONLY_OPS = [
         "conv_bn_relu",
@@ -325,15 +330,15 @@ def get_supported_operator_for_quantization_config(
                 return ops
         return []
 
-    def set_global(self, quantization_config: QuantizationConfig) -> XNNPACKQuantizer:
+    def set_global(self, quantization_config: QuantizationConfig) -> ArmQuantizer:
         self.global_config = quantization_config
         return self
 
     def set_operator_type(
         self,
         operator_type: torch._ops.OpOverloadPacket,
         quantization_config: QuantizationConfig,
-    ) -> XNNPACKQuantizer:
+    ) -> ArmQuantizer:
         self.operator_type_config[operator_type] = quantization_config
         return self
 

backends/arm/arm_quantizer_utils.py

@@ -1,9 +1,14 @@
 # Copyright (c) Meta Platforms, Inc. and affiliates.
+# Copyright 2024 Arm Limited and/or its affiliates.
 # All rights reserved.
 #
 # This source code is licensed under the BSD-style license found in the
 # LICENSE file in the root directory of this source tree.
 
+#
+# Utility functions for ArmQuantizer
+#
+
 import itertools
 import operator
 from dataclasses import dataclass
@@ -62,9 +67,7 @@ class QuantizationConfig:
 
 
 OperatorPatternType = List[Callable]
-OperatorPatternType.__module__ = (
-    "torch.ao.quantization.quantizer.xnnpack_quantizer_utils"
-)
+OperatorPatternType.__module__ = "executorch.backends.arm.arm_quantizer_utils"
 
 AnnotatorType = Callable[
     [
@@ -604,7 +607,7 @@ def _annotate_max_pool2d(
                 maxpool_node = n
         assert (
             maxpool_node is not None
-        ), "XNNPACKQuantizer only works with torch.ops.aten.max_pool2d.default, "
+        ), "ArmQuantizer only works with torch.ops.aten.max_pool2d.default, "
         "please make sure you are exporting the model correctly"
         if _is_annotated([output_node, maxpool_node]):  # type: ignore[list-item]
             continue

backends/arm/test/test_tosa.py

@@ -25,15 +25,14 @@
     _check_ir_validity=False,
 )
 
-from executorch.exir import EdgeCompileConfig
-from executorch.exir.program import to_edge
-from torch.ao.quantization.quantize_pt2e import convert_pt2e, prepare_pt2e
-
 ## For quantization
-from torch.ao.quantization.quantizer.xnnpack_quantizer import (
+from executorch.backends.arm.arm_quantizer import (
+    ArmQuantizer,
     get_symmetric_quantization_config,
-    XNNPACKQuantizer,
 )
+from executorch.exir import EdgeCompileConfig
+from executorch.exir.program import to_edge
+from torch.ao.quantization.quantize_pt2e import convert_pt2e, prepare_pt2e
 
 
 class TestBasicNN(unittest.TestCase):
@@ -88,7 +87,7 @@ def prepare_model_and_ref(test_model, profile=TosaProfile.MI):
             model, copy.deepcopy(model.inputs[profile])
         )
         # Setup the quantizer
-        quantizer = XNNPACKQuantizer()
+        quantizer = ArmQuantizer()
         operator_config = get_symmetric_quantization_config(is_per_channel=False)
         quantizer.set_global(operator_config)
 

backends/arm/test/tester/arm_tester.py

@@ -13,6 +13,10 @@
 )
 
 from executorch.backends.arm.arm_partitioner import ArmPartitioner
+from executorch.backends.arm.arm_quantizer import (
+    ArmQuantizer,
+    get_symmetric_quantization_config,
+)
 
 from executorch.backends.arm.test.tosautil.tosa_test_utils import (
     QuantizationParams,
@@ -29,10 +33,6 @@
 )
 
 from executorch.exir import EdgeCompileConfig
-from torch.ao.quantization.quantizer.xnnpack_quantizer import (
-    get_symmetric_quantization_config,
-    XNNPACKQuantizer,
-)
 from torch.export import ExportedProgram
 
 
@@ -88,9 +88,8 @@ def __init__(
 
     def quantize(self, quantize_stage: Optional[Quantize] = None):
         if quantize_stage is None:
-            # Using the XNNPACKQuantizer for now
             quantize_stage = Quantize(
-                XNNPACKQuantizer(),
+                ArmQuantizer(),
                 get_symmetric_quantization_config(is_per_channel=False),
             )
         return super().quantize(quantize_stage)

examples/arm/aot_arm_compiler.py

@@ -23,21 +23,20 @@
 FORMAT = "[%(levelname)s %(asctime)s %(filename)s:%(lineno)s] %(message)s"
 logging.basicConfig(level=logging.WARNING, format=FORMAT)
 
-# Quantize model if required using the standard export quantizaion flow.
-# For now we're using the xnnpack quantizer as this produces reasonable
-# output for our arithmetic behaviour.
-from torch.ao.quantization.quantize_pt2e import convert_pt2e, prepare_pt2e
-from torch.ao.quantization.quantizer.xnnpack_quantizer import (
+from executorch.backends.arm.arm_quantizer import (
+    ArmQuantizer,
     get_symmetric_quantization_config,
-    XNNPACKQuantizer,
 )
 
+# Quantize model if required using the standard export quantizaion flow.
+from torch.ao.quantization.quantize_pt2e import convert_pt2e, prepare_pt2e
+
 
 def quantize(model, example_inputs):
     """This is the official recommended flow for quantization in pytorch 2.0 export"""
     logging.info("Quantizing Model...")
     logging.debug(f"Original model: {model}")
-    quantizer = XNNPACKQuantizer()
+    quantizer = ArmQuantizer()
     # if we set is_per_channel to True, we also need to add out_variant of quantize_per_channel/dequantize_per_channel
     operator_config = get_symmetric_quantization_config(is_per_channel=False)
     quantizer.set_global(operator_config)