extract sample_ratio for automlx explainability

ads/aqua/constants.py

@@ -43,6 +43,8 @@
 HF_LOGIN_DEFAULT_TIMEOUT = 2
 MODEL_NAME_DELIMITER = ";"
 AQUA_TROUBLESHOOTING_LINK = "https://github.com/oracle-samples/oci-data-science-ai-samples/blob/main/ai-quick-actions/troubleshooting-tips.md"
+MODEL_FILE_DESCRIPTION_VERSION = "1.0"
+MODEL_FILE_DESCRIPTION_TYPE = "modelOSSReferenceDescription"
 
 TRAINING_METRICS_FINAL = "training_metrics_final"
 VALIDATION_METRICS_FINAL = "validation_metrics_final"

ads/aqua/model/entities.py

@@ -15,12 +15,19 @@
 
 import oci
 from huggingface_hub import hf_api
-from pydantic import BaseModel
+from pydantic import BaseModel, Field
+from pydantic.alias_generators import to_camel
 
 from ads.aqua import logger
 from ads.aqua.app import CLIBuilderMixin
 from ads.aqua.common import utils
-from ads.aqua.constants import LIFECYCLE_DETAILS_MISSING_JOBRUN, UNKNOWN_VALUE
+from ads.aqua.config.utils.serializer import Serializable
+from ads.aqua.constants import (
+    LIFECYCLE_DETAILS_MISSING_JOBRUN,
+    MODEL_FILE_DESCRIPTION_TYPE,
+    MODEL_FILE_DESCRIPTION_VERSION,
+    UNKNOWN_VALUE,
+)
 from ads.aqua.data import AquaResourceIdentifier
 from ads.aqua.model.enums import FineTuningDefinedMetadata
 from ads.aqua.training.exceptions import exit_code_dict
@@ -304,3 +311,75 @@ class ImportModelDetails(CLIBuilderMixin):
 
     def __post_init__(self):
         self._command = "model register"
+
+
+class ModelFileInfo(Serializable):
+    """Describes the file information of this model.
+
+    Attributes:
+        name (str): The name of the model artifact file.
+        version (str): The version of the model artifact file.
+        size_in_bytes (int): The size of the model artifact file in bytes.
+    """
+
+    name: str = Field(..., description="The name of model artifact file.")
+    version: str = Field(..., description="The version of model artifact file.")
+    size_in_bytes: int = Field(
+        ..., description="The size of model artifact file in bytes."
+    )
+
+    class Config:
+        alias_generator = to_camel
+        extra = "allow"
+
+
+class ModelArtifactInfo(Serializable):
+    """Describes the artifact information of this model.
+
+    Attributes:
+        namespace (str): The namespace of the model artifact location.
+        bucket_name (str): The bucket name of model artifact location.
+        prefix (str): The prefix of model artifact location.
+        objects: (List[ModelFileInfo]): A list of model artifact objects.
+    """
+
+    namespace: str = Field(
+        ..., description="The name space of model artifact location."
+    )
+    bucket_name: str = Field(
+        ..., description="The bucket name of model artifact location."
+    )
+    prefix: str = Field(..., description="The prefix of model artifact location.")
+    objects: List[ModelFileInfo] = Field(
+        ..., description="List of model artifact objects."
+    )
+
+    class Config:
+        alias_generator = to_camel
+        extra = "allow"
+
+
+class ModelFileDescription(Serializable):
+    """Describes the model file description.
+
+    Attributes:
+        version (str): The version of the model file description. Defaults to `1.0`.
+        type (str): The type of model file description. Defaults to `modelOSSReferenceDescription`.
+        models List[ModelArtifactInfo]: A list of model artifact information.
+    """
+
+    version: str = Field(
+        default=MODEL_FILE_DESCRIPTION_VERSION,
+        description="The version of model file description.",
+    )
+    type: str = Field(
+        default=MODEL_FILE_DESCRIPTION_TYPE,
+        description="The type of model file description.",
+    )
+    models: List[ModelArtifactInfo] = Field(
+        ..., description="List of model artifact information."
+    )
+
+    class Config:
+        alias_generator = to_camel
+        extra = "allow"

ads/aqua/model/model.py

@@ -79,6 +79,7 @@
     AquaModelReadme,
     AquaModelSummary,
     ImportModelDetails,
+    ModelFileDescription,
     ModelValidationResult,
 )
 from ads.aqua.model.enums import MultiModelSupportedTaskType
@@ -271,8 +272,8 @@ def create_multi(
                 "Model list cannot be empty. Please provide at least one model for deployment."
             )
 
-        artifact_list = []
         display_name_list = []
+        model_file_description_list: List[ModelFileDescription] = []
         model_custom_metadata = ModelCustomMetadata()
 
         service_inference_containers = (
@@ -299,6 +300,7 @@ def create_multi(
         for model in models:
             source_model = DataScienceModel.from_id(model.model_id)
             display_name = source_model.display_name
+            model_file_description = source_model.model_file_description
             # Update model name in user's input model
             model.model_name = model.model_name or display_name
 
@@ -324,7 +326,15 @@ def create_multi(
             # Update model artifact location in user's input model
             model.artifact_location = model_artifact_path
 
-            artifact_list.append(model_artifact_path)
+            if not model_file_description:
+                raise AquaValueError(
+                    f"Model '{display_name}' (ID: {model.model_id}) has no file description. "
+                    "Please register the model first."
+                )
+
+            model_file_description_list.append(
+                ModelFileDescription(**model_file_description)
+            )
 
             # Validate deployment container consistency
             deployment_container = source_model.custom_metadata_list.get(
@@ -402,9 +412,16 @@ def create_multi(
             .with_custom_metadata_list(model_custom_metadata)
         )
 
-        # Attach artifacts
-        for artifact in artifact_list:
-            custom_model.add_artifact(uri=artifact)
+        # Update multi model file description to attach artifacts
+        custom_model.with_model_file_description(
+            json_dict=ModelFileDescription(
+                models=[
+                    models
+                    for model_file_description in model_file_description_list
+                    for models in model_file_description.models
+                ]
+            ).model_dump(by_alias=True)
+        )
 
         # Finalize creation
         custom_model.create(model_by_reference=True)

tests/unitary/with_extras/aqua/test_model.py

@@ -225,6 +225,35 @@ class TestDataset:
         },
     ]
 
+    model_file_description = {
+        "version": "1.0",
+        "type": "modelOSSReferenceDescription",
+        "models": [
+            {
+                "namespace": "test_namespace",
+                "bucketName": "test_bucket",
+                "prefix": "models/meta-llama/Llama-3.2-3B-Instruct",
+                "objects": [
+                    {
+                        "name": "models/meta-llama/Llama-3.2-3B-Instruct/.gitattributes",
+                        "version": "bfbf278c-10af-4f2c-8240-11fed02e1322",
+                        "sizeInBytes": 1519,
+                    },
+                    {
+                        "name": "models/meta-llama/Llama-3.2-3B-Instruct/LICENSE.txt",
+                        "version": "4238d1e2-d826-4300-a344-0ead410afa27",
+                        "sizeInBytes": 7712,
+                    },
+                    {
+                        "name": "models/meta-llama/Llama-3.2-3B-Instruct/README.md",
+                        "version": "57382552-9ad0-4546-b38c-c96634f3b8a2",
+                        "sizeInBytes": 41744,
+                    },
+                ],
+            }
+        ],
+    }
+
     SERVICE_COMPARTMENT_ID = "ocid1.compartment.oc1..<OCID>"
     COMPARTMENT_ID = "ocid1.compartment.oc1..<UNIQUE_OCID>"
     SERVICE_MODEL_ID = "ocid1.datasciencemodel.oc1.iad.<OCID>"
@@ -360,24 +389,20 @@ def test_create_model(self, mock_from_id, mock_validate, mock_create):
         )
         assert model.provenance_metadata.training_id == "test_training_id"
 
-    @patch.object(DataScienceModel, "add_artifact")
     @patch.object(DataScienceModel, "create_custom_metadata_artifact")
     @patch.object(DataScienceModel, "create")
-    @patch("ads.model.datascience_model.validate")
     @patch.object(AquaApp, "get_container_config")
     @patch.object(DataScienceModel, "from_id")
     def test_create_multimodel(
         self,
         mock_from_id,
         mock_get_container_config,
-        mock_validate,
         mock_create,
         mock_create_custom_metadata_artifact,
-        mock_add_artifact,
     ):
         mock_get_container_config.return_value = get_container_config()
         mock_model = MagicMock()
-        mock_model.model_file_description = {"test_key": "test_value"}
+        mock_model.model_file_description = TestDataset.model_file_description
         mock_model.display_name = "test_display_name"
         mock_model.description = "test_description"
         mock_model.freeform_tags = {
@@ -396,14 +421,14 @@ def test_create_multimodel(
         model_info_1 = AquaMultiModelRef(
             model_id="test_model_id_1",
             gpu_count=2,
-            model_task = "text_embedding",
+            model_task="text_embedding",
             env_var={"params": "--trust-remote-code --max-model-len 60000"},
         )
 
         model_info_2 = AquaMultiModelRef(
             model_id="test_model_id_2",
             gpu_count=2,
-            model_task = "image_text_to_text",
+            model_task="image_text_to_text",
             env_var={"params": "--trust-remote-code --max-model-len 32000"},
         )
 
@@ -455,10 +480,10 @@ def test_create_multimodel(
         mock_model.freeform_tags["task"] = "unsupported_task"
         with pytest.raises(AquaValueError):
             model = self.app.create_multi(
-                    models=[model_info_1, model_info_2],
-                    project_id="test_project_id",
-                    compartment_id="test_compartment_id",
-                )
+                models=[model_info_1, model_info_2],
+                project_id="test_project_id",
+                compartment_id="test_compartment_id",
+            )
 
         mock_model.freeform_tags["task"] = "text-generation"
         model_info_1.model_task = "text_embedding"
@@ -470,9 +495,7 @@ def test_create_multimodel(
             compartment_id="test_compartment_id",
         )
 
-        mock_add_artifact.assert_called()
         mock_from_id.assert_called()
-        mock_validate.assert_not_called()
         mock_create.assert_called_with(model_by_reference=True)
 
         mock_model.compartment_id = TestDataset.SERVICE_COMPARTMENT_ID