TRN2 Meshes and Configurations

Author: apoorvtintin

axlearn/common/trainer_config_modifier.py

@@ -2,12 +2,13 @@
 
 """Defines trainer config modifiers, which will be used in model definitions."""
 
-from typing import Dict, Sequence, Union
+from typing import Callable, Dict, NamedTuple, Sequence, Union
 
 from axlearn.common import config
 from axlearn.common.base_layer import RematSpec
 from axlearn.common.config import (
     REQUIRED,
+    ConfigBase,
     ConfigModifier,
     ConfigOr,
     Required,
@@ -17,7 +18,53 @@
 from axlearn.common.gradient_accumulation import with_minibatch_steps
 from axlearn.common.metrics import MetricAccumulator
 from axlearn.common.trainer import SpmdTrainer
-from axlearn.common.utils import HybridMeshShape, MeshShape
+from axlearn.common.utils import HybridMeshShape, MeshShape, PartitionSpec
+
+
+class _FoundModule(NamedTuple):
+    """Module found in recursive search of matching module."""
+
+    # The module found
+    module: ConfigModifier.Config
+    # The parent of the module found
+    parent_module: ConfigModifier.Config
+    # Key of the found module in parent
+    key_in_parent: ConfigModifier.Config
+
+
+def _find_target_module(module_name: str, cfg: SpmdTrainer.Config) -> _FoundModule:
+    """Recursively search for the target module matching module_name in provided cfg.
+
+    Args:
+        module_name: Name of the target module
+        cfg: The trainer config to be searched for module_name
+
+    Raises:
+        ValueError: The module_name is not found.
+
+    Returns:
+        A Tuple(curr_module, key_in_parent, parent_module)
+            curr_module: Module found
+            parent_module: The parent module
+            key_in_parent: Key in parent for the found module
+    """
+
+    # Here we assume x.y.z format.
+    # One example would be model.decoder.transformer.layer.
+    target_modules = module_name.split(".")
+    curr_module = cfg
+    key_in_parent = None
+    parent_module = None
+
+    for target_module in target_modules:
+        if not hasattr(curr_module, target_module):
+            raise ValueError(f"{target_module} is not found in {curr_module}.")
+        parent_module = curr_module
+        key_in_parent = target_module
+        curr_module = getattr(curr_module, target_module)
+    return _FoundModule(
+        module=curr_module, parent_module=parent_module, key_in_parent=key_in_parent
+    )
 
 
 class GradientAccumulationModifier(ConfigModifier):
@@ -100,18 +147,11 @@ def __call__(self, cfg: SpmdTrainer.Config) -> SpmdTrainer.Config:
         """
 
         for module_name, remat_spec in self._remat_policies.items():
-            # Here we assume x.y.z format.
-            # One example would be model.decoder.transformer.layer.
-            target_modules = module_name.split(".")
-            curr_module = cfg
-            for target_module in target_modules:
-                if not hasattr(curr_module, target_module):
-                    raise ValueError(f"{target_module} is not found in {curr_module}.")
-                curr_module = getattr(curr_module, target_module)
+            found_module = _find_target_module(module_name, cfg)
             # Here we assume all modules have remat_spec attribute.
-            if not hasattr(curr_module, "remat_spec"):
-                raise ValueError(f"{curr_module} does not have remat_spec attribute")
-            curr_module.remat_spec = remat_spec
+            if not hasattr(found_module.module, "remat_spec"):
+                raise ValueError(f"{found_module.module} does not have remat_spec attribute")
+            found_module.module.remat_spec = remat_spec
         return cfg
 
 
@@ -146,6 +186,113 @@ def __call__(self, cfg: SpmdTrainer.Config) -> SpmdTrainer.Config:
         return cfg
 
 
+class ModelConfigModifier(ConfigModifier):
+    """Update the model config for the trainer config."""
+
+    @config_class
+    class Config(ConfigModifier.Config):
+        """Configure ModelConfigModifier.
+
+        Attributes:
+            model_cfg_modifications: A mapping from module path
+                (e.g. `model.decoder.transformer.layer`) to a Config.
+        """
+
+        model_cfg_modifications: Dict[str, Callable[[ConfigBase], ConfigBase]] = {}
+
+    def __init__(self, cfg: Config):
+        super().__init__(cfg)
+        self._model_cfg_modifications = self.config.model_cfg_modifications
+
+    def _merge_configs(self, target_cfg: ConfigBase, found_module: ConfigBase) -> ConfigBase:
+        """Merge configurations from the config being replaced on a best effort basis.
+
+        Args:
+            target_cfg: configuration that will replace found_module.
+            found_module: existing configuration whose class will be replaced
+                but it's confguration will be merged with target_cfg.
+
+        Returns:
+            The modified config.
+
+        """
+        for key in target_cfg.keys():
+            if key == "klass":
+                continue
+            elif hasattr(found_module.module, key) and hasattr(target_cfg, key):
+                setattr(target_cfg, key, getattr(found_module.module, key))
+        return target_cfg
+
+    def __call__(self, cfg: SpmdTrainer.Config) -> SpmdTrainer.Config:
+        """Overwrite the model config of the specified modules.
+
+        Args:
+            cfg: The trainer config to be modified.
+
+        Raises:
+            ValueError: The target module is not found.
+
+        Returns:
+            The modified trainer config.
+        """
+
+        # Iterate over modules in the mapping, modules are sorted based on module name length.
+        # This ensures parent is modified before children to avoid missing modifications.
+        for module_name, model_cfg in sorted(
+            self._model_cfg_modifications.items(), key=lambda item: len(item[0])
+        ):
+            found_module = _find_target_module(module_name, cfg)
+
+            model_cfg = self._merge_configs(model_cfg, found_module)
+            # Replace in the parent config
+            setattr(found_module.parent_module, found_module.key_in_parent, model_cfg)
+        return cfg
+
+
+class PartitionSpecModifier(ConfigModifier):
+    """Update the partition spec attribute for the specified modules."""
+
+    @config_class
+    class Config(ConfigModifier.Config):
+        """Configure PartitionSpecModifier.
+
+        Attributes:
+            partition_specs: A nested mapping from module path
+                (e.g. `model.decoder.transformer.layer`) to another
+                mapping of model attribute to PartitionSpec.
+        """
+
+        partition_specs: Required[Dict[str, PartitionSpec]] = REQUIRED
+
+    def __init__(self, cfg: Config):
+        super().__init__(cfg)
+        self._attribute_dicts = self.config.partition_specs
+
+    def __call__(self, cfg: SpmdTrainer.Config) -> SpmdTrainer.Config:
+        """Update the partition_spec attributes for the specified modules.
+
+        Args:
+            cfg: The trainer config to be modified.
+
+        Raises:
+            ValueError: The target module is not found.
+            ValueError: The partition_spec attribute is not found.
+
+        Returns:
+            The modified trainer config.
+        """
+        for module_name, partition_spec_dict in self._attribute_dicts.items():
+            found_module = _find_target_module(module_name, cfg)
+            for partition_spec_name, partition_spec in partition_spec_dict.items():
+                if not hasattr(found_module.module, partition_spec_name):
+                    raise ValueError(
+                        f"{found_module.module} does not have {partition_spec_name} attribute"
+                    )
+                setattr(found_module.module, partition_spec_name, partition_spec)
+
+        return cfg
+
+
 class ChainConfigModifier(ConfigModifier):
     """Chain multiple config modifiers together."""
 

axlearn/common/trainer_config_modifier_test.py

@@ -5,13 +5,16 @@
 import jax
 from absl.testing import absltest
 
-from axlearn.common import test_utils
+from axlearn.common import causal_lm, test_utils
+from axlearn.common.attention import RepeatedTransformerLayer, StackedTransformerLayer
 from axlearn.common.base_layer import RematSpec
 from axlearn.common.trainer import SpmdTrainer
 from axlearn.common.trainer_config_modifier import (
     ChainConfigModifier,
     GradientAccumulationModifier,
     MeshShapeModifier,
+    ModelConfigModifier,
+    PartitionSpecModifier,
     RematSpecModifier,
 )
 from axlearn.common.trainer_test import DummyModel
@@ -65,6 +68,88 @@ def test_remat_policy_override(self):
             _ = cfg_modifier(cfg)
 
 
+class ModelConfigModifierTest(test_utils.TestCase):
+    def test_model_config_override(self):
+        cfg = SpmdTrainer.default_config().set(model=causal_lm.Model.default_config())
+        self.assertTrue(
+            str(cfg.model.decoder.transformer) == str(StackedTransformerLayer.default_config())
+        )
+
+        cfg_modifier = (
+            ModelConfigModifier.default_config()
+            .set(
+                model_cfg_modifications={
+                    "model.decoder.transformer": RepeatedTransformerLayer.default_config(),
+                }
+            )
+            .instantiate()
+        )
+
+        cfg = cfg_modifier(cfg)
+        # The default StackedTransformerLayer should have changed to RepeatedTransformerLayer
+        self.assertTrue(
+            str(cfg.model.decoder.transformer) == str(RepeatedTransformerLayer.default_config())
+        )
+        cfg_modifier = (
+            ModelConfigModifier.default_config()
+            .set(
+                model_cfg_modifications={
+                    "model.decoder.unknown": RepeatedTransformerLayer.default_config(),
+                }
+            )
+            .instantiate()
+        )
+        # Ensure that the exception is working.
+        with self.assertRaisesRegex(ValueError, "unknown is not found in.*"):
+            _ = cfg_modifier(cfg)
+
+
+class PartitionSpecModifierTest(test_utils.TestCase):
+    def test_partition_spec_override(self):
+        cfg = SpmdTrainer.default_config().set(model=DummyModel.default_config())
+        cfg_modifier = (
+            PartitionSpecModifier.default_config()
+            .set(
+                partition_specs={
+                    "model.linear": {"param_partition_spec": ("model", ("expert", "fsdp", "seq"))},
+                },
+            )
+            .instantiate()
+        )
+        cfg = cfg_modifier(cfg)
+        self.assertTrue(
+            str(cfg.model.linear.param_partition_spec), """("model", ("expert", "fsdp", "seq")"""
+        )
+        cfg_modifier = (
+            PartitionSpecModifier.default_config()
+            .set(
+                partition_specs={
+                    "model.linear": {"param_partition_spec": ("model", ("expert", "fsdp", "seq"))},
+                    "model.unknown": {"param_partition_spec": ("model", ("expert", "fsdp", "seq"))},
+                },
+            )
+            .instantiate()
+        )
+        # Ensure that the exception is working.
+        with self.assertRaisesRegex(ValueError, "unknown is not found in.*"):
+            _ = cfg_modifier(cfg)
+
+        cfg_modifier = (
+            PartitionSpecModifier.default_config()
+            .set(
+                partition_specs={
+                    "model.linear": {
+                        "param_partition_spec": ("model", ("expert", "fsdp", "seq")),
+                        "unknown_partition_spec": ("model", ("expert", "fsdp", "seq")),
+                    },
+                },
+            )
+            .instantiate()
+        )
+        with self.assertRaisesRegex(ValueError, ".*does not have unknown_partition_spec attribute"):
+            _ = cfg_modifier(cfg)
+
+
 class MeshShapeModifierTest(test_utils.TestCase):
     def test_mesh_shape_update(self):
         cfg = SpmdTrainer.default_config().set(model=DummyModel.default_config())