[Executorch][Export/ Recipes] Modify pre_edge_transform_passes, edge_transform_passes definition to take exported program and method name. (#14178) (D81730890)

Co-authored-by: Abhinay Kukkadapu <[email protected]>

Differential version: D81730890

Author: pytorchbot

export/export.py

@@ -195,12 +195,12 @@ def _build_stages(self, stages: List[StageType]) -> Dict[StageType, Stage]:
             elif stage_type == StageType.QUANTIZE:
                 stage = QuantizeStage(self._quant_recipe)
             elif stage_type == StageType.TORCH_EXPORT:
-                pre_edge_passes = None
-                if self._export_recipe.pre_edge_transform_passes is not None:
-                    pre_edge_passes = list(
-                        self._export_recipe.pre_edge_transform_passes
+                aten_transform_passes = None
+                if self._export_recipe.aten_transform_passes is not None:
+                    aten_transform_passes = list(
+                        self._export_recipe.aten_transform_passes
                     )
-                stage = TorchExportStage(pre_edge_passes)
+                stage = TorchExportStage(aten_transform_passes)
             elif stage_type == StageType.TO_EDGE_TRANSFORM_AND_LOWER:
                 stage = EdgeTransformAndLowerStage.from_recipe(self._lowering_recipe)
             elif stage_type == StageType.TO_EDGE:

export/recipe.py

@@ -7,9 +7,10 @@
 from abc import ABCMeta, abstractmethod
 from dataclasses import dataclass
 from enum import Enum, EnumMeta
-from typing import Callable, List, Optional, Sequence
+from typing import Callable, List, Optional
 
 import torch
+from executorch.exir import ExportedProgram
 
 from executorch.exir._warnings import experimental
 
@@ -117,12 +118,15 @@ class LoweringRecipe:
 
     Attributes:
         partitioners: Optional list of partitioners for model partitioning
-        edge_transform_passes: Optional sequence of transformation passes to apply
+        edge_transform_passes: Optional list of callables that take (method_name: str, exported_program: ExportedProgram) as arguments
+                               and return a list of passes (PassType) to be executed during lowering stages.
         edge_compile_config: Optional edge compilation configuration
     """
 
     partitioners: Optional[List[Partitioner]] = None
-    edge_transform_passes: Optional[Sequence[PassType]] = None
+    edge_transform_passes: (
+        None | List[Callable[[str, ExportedProgram], List[PassType]]]
+    ) = None
     # pyre-ignore[11]: Type not defined
     edge_compile_config: Optional[EdgeCompileConfig] = None
 
@@ -141,8 +145,8 @@ class ExportRecipe:
     Attributes:
         name: Optional name for the recipe
         quantization_recipe: Optional quantization recipe for model quantization
-        pre_edge_transform_passes: Optional function to apply transformation passes
-                                  before edge lowering
+        aten_transform_passes: Optional list of functions to apply transformation passes to the program before edge lowering.
+                               These callables are invoked to modify and return the transformed program.
         lowering_recipe: Optional lowering recipe for model lowering and partitioning
         executorch_backend_config: Optional backend configuration for ExecuTorch
         pipeline_stages: Optional list of stages to execute, defaults to a standard pipeline.
@@ -151,7 +155,9 @@ class ExportRecipe:
 
     name: Optional[str] = None
     quantization_recipe: Optional[QuantizationRecipe] = None
-    pre_edge_transform_passes: Optional[Sequence[PassType]] = None
+    aten_transform_passes: Optional[
+        List[Callable[[str, ExportedProgram], ExportedProgram]]
+    ] = None
     lowering_recipe: Optional[LoweringRecipe] = None
     # pyre-ignore[11]: Type not defined
     executorch_backend_config: Optional[ExecutorchBackendConfig] = None
@@ -240,8 +246,8 @@ def _combine_recipes(  # noqa: C901
 
         for recipe in backend_recipes:
             # Collect pre-edge transform passes
-            if recipe.pre_edge_transform_passes:
-                all_pre_edge_passes.extend(recipe.pre_edge_transform_passes)
+            if recipe.aten_transform_passes:
+                all_pre_edge_passes.extend(recipe.aten_transform_passes)
 
             # Collect partitioners from lowering recipes
             if recipe.lowering_recipe and recipe.lowering_recipe.partitioners:
@@ -307,7 +313,7 @@ def _combine_recipes(  # noqa: C901
         return cls(
             name=recipe_name,
             quantization_recipe=combined_quantization_recipe,
-            pre_edge_transform_passes=all_pre_edge_passes,
+            aten_transform_passes=all_pre_edge_passes,
             lowering_recipe=combined_lowering_recipe,
             executorch_backend_config=combined_backend_config,
         )

export/stages.py

@@ -7,14 +7,14 @@
 import copy
 import logging
 from abc import ABC, abstractmethod
-from typing import Any, Callable, Dict, List, Optional, Sequence
+from collections import defaultdict
+from typing import Any, Callable, Dict, List, Optional
 
 import torch
 from executorch.devtools.backend_debug import get_delegation_info
-from executorch.exir import EdgeCompileConfig
+from executorch.exir import EdgeCompileConfig, ExportedProgram
 from executorch.exir.backend.backend_api import validation_disabled
 from executorch.exir.program import to_edge, to_edge_transform_and_lower
-from executorch.exir.program._program import _transform
 from executorch.export.recipe import LoweringRecipe, QuantizationRecipe
 from executorch.export.types import StageType
 from torch import nn
@@ -107,10 +107,12 @@ class TorchExportStage(Stage):
 
     def __init__(
         self,
-        pre_edge_transform_passes: Optional[List[PassType]] = None,
+        aten_transform_passes: Optional[
+            List[Callable[[str, ExportedProgram], ExportedProgram]]
+        ] = None,
     ) -> None:
         super().__init__()
-        self._pre_edge_transform_passes = pre_edge_transform_passes
+        self._aten_transform_passes = aten_transform_passes
 
     @property
     def stage_type(self) -> str:
@@ -149,9 +151,13 @@ def run(self, artifact: PipelineArtifact) -> None:
                 )
 
                 # Apply pre-edge transform passes if available
-                for pass_ in self._pre_edge_transform_passes or []:
-                    exported_programs[method_name] = _transform(
-                        exported_programs[method_name], pass_
+                for pass_ in self._aten_transform_passes or []:
+                    if not callable(pass_):
+                        raise ValueError(
+                            "Aten transform passes must be a callable that can transform and return an exported program"
+                        )
+                    exported_programs[method_name] = pass_(
+                        method_name, exported_programs[method_name]
                     )
 
         self._artifact = artifact.copy_with_new_data(exported_programs)
@@ -165,7 +171,9 @@ class EdgeTransformAndLowerStage(Stage):
     def __init__(
         self,
         partitioners: Optional[List[Any]] = None,
-        transform_passes: Optional[Sequence[Callable[[Any], Optional[Any]]]] = None,
+        transform_passes: (
+            None | List[Callable[[str, ExportedProgram], List[PassType]]]
+        ) = None,
         compile_config: Optional[Any] = None,
     ) -> None:
         self._partitioners = partitioners
@@ -205,11 +213,28 @@ def run(self, artifact: PipelineArtifact) -> None:
         constant_methods = artifact.get_context("constant_methods")
         generate_etrecord = artifact.get_context("generate_etrecord", False)
 
+        # per method transform passes
+        transform_passes = defaultdict(list)
+        for method_name, ep in exported_programs.items():
+            # Resolve transform passes from callable
+            for pass_ in self._transform_passes or []:
+                if not callable(pass_):
+                    raise ValueError(
+                        "Transform passes must be a callable that resolves to a list of passes"
+                    )
+                passes = pass_(method_name, ep)
+                if isinstance(passes, list):
+                    transform_passes[method_name].extend(passes)
+                else:
+                    raise ValueError(
+                        "Transform passes must be a callable that resolves to a list of passes"
+                    )
+
         with validation_disabled():
             edge_program_manager = to_edge_transform_and_lower(
                 exported_programs,
                 partitioner=self._partitioners,
-                transform_passes=self._transform_passes,
+                transform_passes=transform_passes,
                 constant_methods=constant_methods,
                 compile_config=self._compile_config,
                 generate_etrecord=generate_etrecord,
@@ -396,7 +421,7 @@ def run(self, artifact: PipelineArtifact) -> None:
             captured_graph = torch.export.export(model, inputs, strict=True).module()
 
             quantizer = self._get_quantizer_for_prepare_pt2e(
-                self._quantization_recipe.quantizers
+                self._quantization_recipe.quantizers  # pyre-ignore
             )
             prepared_model = prepare_pt2e(captured_graph, quantizer)
 
@@ -471,7 +496,9 @@ class ToBackendStage(Stage):
     def __init__(
         self,
         partitioners: Optional[List[Any]] = None,
-        transform_passes: Optional[Sequence[Callable[[Any], Optional[Any]]]] = None,
+        transform_passes: (
+            None | List[Callable[[str, ExportedProgram], List[PassType]]]
+        ) = None,
     ) -> None:
         super().__init__()
         self._partitioners = partitioners
@@ -513,11 +540,24 @@ def run(self, artifact: PipelineArtifact) -> None:
         if edge_program_manager is None:
             raise RuntimeError("Edge program manager is not set.")
 
-        # Apply transform passes if available
-        if self._transform_passes:
-            edge_program_manager = edge_program_manager.transform(
-                self._transform_passes
-            )
+        # per method transform passes
+        transform_passes = defaultdict(list)
+        for method_name in edge_program_manager.methods:
+            # Resolve transform passes if it's a callable
+            ep = edge_program_manager.exported_program(method_name)
+            for pass_ in self._transform_passes or []:
+                if not callable(pass_):
+                    raise ValueError(
+                        "Transform passes must be a callable that resolves to a list of passes"
+                    )
+                passes = pass_(method_name, ep)
+                if isinstance(passes, list):
+                    transform_passes[method_name].extend(passes)
+                else:
+                    raise ValueError("Transform passes must return list of passes")
+
+        # Apply transform passes
+        edge_program_manager = edge_program_manager.transform(transform_passes)
 
         # Apply partitioners if available
         if self._partitioners is not None and len(self._partitioners) > 0: