improve validation

Author: VipulMascarenhas

ads/aqua/finetuning/constants.py

@@ -16,7 +16,7 @@ class FineTuneCustomMetadata(str, metaclass=ExtendedEnumMeta):
     SERVICE_MODEL_FINE_TUNE_CONTAINER = "finetune-container"
 
 
-class FineTuningForbiddenParams(str, metaclass=ExtendedEnumMeta):
+class FineTuningRestrictedParams(str, metaclass=ExtendedEnumMeta):
     OPTIMIZER = "optimizer"
 
 

ads/aqua/finetuning/entities.py

@@ -3,22 +3,22 @@
 # Licensed under the Universal Permissive License v 1.0 as shown at https://oss.oracle.com/licenses/upl/
 
 import json
-from typing import List, Optional
+from typing import List, Literal, Optional, Union
 
 from pydantic import Field, model_validator
 
 from ads.aqua.common.errors import AquaValueError
 from ads.aqua.config.utils.serializer import Serializable
 from ads.aqua.data import AquaResourceIdentifier
-from ads.aqua.finetuning.constants import FineTuningForbiddenParams
+from ads.aqua.finetuning.constants import FineTuningRestrictedParams
 
 
 class AquaFineTuningParams(Serializable):
     """Class for maintaining aqua fine-tuning model parameters"""
 
     epochs: Optional[int] = None
     learning_rate: Optional[float] = None
-    sample_packing: Optional[bool] = "auto"
+    sample_packing: Union[bool, None, Literal["auto"]] = "auto"
     batch_size: Optional[int] = (
         None  # make it batch_size for user, but internally this is micro_batch_size
     )
@@ -40,16 +40,18 @@ def to_dict(self) -> dict:
 
     @model_validator(mode="before")
     @classmethod
-    def validate_forbidden_fields(cls, data: dict):
+    def validate_restricted_fields(cls, data: dict):
         # we may want to skip validation if loading data from config files instead of user entered parameters
         validate = data.pop("_validate", True)
         if not (validate and isinstance(data, dict)):
             return data
-        forbidden_params = [
-            param for param in data if param in FineTuningForbiddenParams.values()
+        restricted_params = [
+            param for param in data if param in FineTuningRestrictedParams.values()
         ]
-        if forbidden_params:
-            raise AquaValueError(f"Found restricted parameter name: {forbidden_params}")
+        if restricted_params:
+            raise AquaValueError(
+                f"Found restricted parameter name: {restricted_params}"
+            )
         return data
 
 

ads/aqua/finetuning/finetuning.py

@@ -152,7 +152,7 @@ def create(
                 f"Logging is required for fine tuning if replica is larger than {DEFAULT_FT_REPLICA}."
             )
 
-        ft_parameters = self._validate_finetuning_params(
+        ft_parameters = self._get_finetuning_params(
             create_fine_tuning_details.ft_parameters
         )
 
@@ -591,8 +591,9 @@ def get_finetuning_default_params(self, model_id: str) -> Dict:
         default_params = {"params": {}}
         finetuning_config = self.get_finetuning_config(model_id)
         config_parameters = finetuning_config.get("configuration", UNKNOWN_DICT)
-        config_parameters["_validate"] = False
-        dataclass_fields = AquaFineTuningParams(**config_parameters).to_dict()
+        dataclass_fields = self._get_finetuning_params(
+            config_parameters, validate=False
+        ).to_dict()
         for name, value in config_parameters.items():
             if name in dataclass_fields:
                 if name == "micro_batch_size":
@@ -602,9 +603,17 @@ def get_finetuning_default_params(self, model_id: str) -> Dict:
         return default_params
 
     @staticmethod
-    def _validate_finetuning_params(params: Dict = None) -> AquaFineTuningParams:
+    def _get_finetuning_params(
+        params: Dict = None, validate: bool = True
+    ) -> AquaFineTuningParams:
+        """
+        Get and validate the fine-tuning params, and return an error message if validation fails. In order to skip
+        @model_validator decorator's validation, pass validate=False.
+        """
         try:
-            finetuning_params = AquaFineTuningParams(**params)
+            finetuning_params = AquaFineTuningParams(
+                **{**params, **{"_validate": validate}}
+            )
         except ValidationError as ex:
             # combine both loc and msg for errors where loc (field) is present in error details, else only build error
             # message using msg field. Added to handle error messages from pydantic model validator handler.
@@ -631,5 +640,5 @@ def validate_finetuning_params(self, params: Dict = None) -> Dict:
         -------
             Return a list of restricted params.
         """
-        self._validate_finetuning_params(params or {})
+        self._get_finetuning_params(params or {})
         return {"valid": True}