Merge branch 'main' of https://github.com/oracle/accelerated-data-science into ODSC-63978/onnx_model_support

Author: lu-ohai

THIRD_PARTY_LICENSES.txt

@@ -24,6 +24,12 @@ autots
 * Source code: https://github.com/winedarksea/AutoTS
 * Project home: https://winedarksea.github.io/AutoTS/build/html/index.html
 
+autogen
+* Copyright (c) 2024 Microsoft Corporation.
+* License: MIT License
+* Source code: https://github.com/microsoft/autogen
+* Project home: microsoft.github.io/autogen/
+
 bokeh
 * Copyright Copyright (c) 2012 - 2021, Anaconda, Inc., and Bokeh Contributors
 * License: BSD 3-Clause "New" or "Revised" License

ads/aqua/app.py

@@ -2,6 +2,7 @@
 # Copyright (c) 2024 Oracle and/or its affiliates.
 # Licensed under the Universal Permissive License v 1.0 as shown at https://oss.oracle.com/licenses/upl/
 
+import json
 import os
 from dataclasses import fields
 from typing import Dict, Union
@@ -135,6 +136,8 @@ def create_model_version_set(
         description: str = None,
         compartment_id: str = None,
         project_id: str = None,
+        freeform_tags: dict = None,
+        defined_tags: dict = None,
         **kwargs,
     ) -> tuple:
         """Creates ModelVersionSet from given ID or Name.
@@ -153,7 +156,10 @@ def create_model_version_set(
             Project OCID.
         tag: (str, optional)
             calling tag, can be Tags.AQUA_FINE_TUNING or Tags.AQUA_EVALUATION
-
+        freeform_tags: (dict, optional)
+            Freeform tags for the model version set
+        defined_tags: (dict, optional)
+            Defined tags for the model version set
         Returns
         -------
         tuple: (model_version_set_id, model_version_set_name)
@@ -182,13 +188,15 @@ def create_model_version_set(
                 mvs_freeform_tags = {
                     tag: tag,
                 }
+                mvs_freeform_tags = {**mvs_freeform_tags, **(freeform_tags or {})}
                 model_version_set = (
                     ModelVersionSet()
                     .with_compartment_id(compartment_id)
                     .with_project_id(project_id)
                     .with_name(model_version_set_name)
                     .with_description(description)
                     .with_freeform_tags(**mvs_freeform_tags)
+                    .with_defined_tags(**(defined_tags or {}))
                     # TODO: decide what parameters will be needed
                     # when refactor eval to use this method, we need to pass tag here.
                     .create(**kwargs)
@@ -340,7 +348,9 @@ def build_cli(self) -> str:
         """
         cmd = f"ads aqua {self._command}"
         params = [
-            f"--{field.name} {getattr(self,field.name)}"
+            f"--{field.name} {json.dumps(getattr(self, field.name))}"
+            if isinstance(getattr(self, field.name), dict)
+            else f"--{field.name} {getattr(self, field.name)}"
             for field in fields(self.__class__)
             if getattr(self, field.name) is not None
         ]

ads/aqua/common/utils.py

@@ -788,13 +788,14 @@ def get_ocid_substring(ocid: str, key_len: int) -> str:
     return ocid[-key_len:] if ocid and len(ocid) > key_len else ""
 
 
-def upload_folder(os_path: str, local_dir: str, model_name: str) -> str:
+def upload_folder(os_path: str, local_dir: str, model_name: str, exclude_pattern: str = None) -> str:
     """Upload the local folder to the object storage
 
     Args:
         os_path (str): object storage URI with prefix. This is the path to upload
         local_dir (str): Local directory where the object is downloaded
         model_name (str): Name of the huggingface model
+        exclude_pattern (optional, str): The matching pattern of files to be excluded from uploading.
     Retuns:
         str: Object name inside the bucket
     """
@@ -804,6 +805,8 @@ def upload_folder(os_path: str, local_dir: str, model_name: str) -> str:
     auth_state = AuthState()
     object_path = os_details.filepath.rstrip("/") + "/" + model_name + "/"
     command = f"oci os object bulk-upload --src-dir {local_dir} --prefix {object_path} -bn {os_details.bucket} -ns {os_details.namespace} --auth {auth_state.oci_iam_type} --profile {auth_state.oci_key_profile} --no-overwrite"
+    if exclude_pattern:
+        command += f" --exclude {exclude_pattern}"
     try:
         logger.info(f"Running: {command}")
         subprocess.check_call(shlex.split(command))

ads/aqua/constants.py

@@ -35,6 +35,7 @@
 AQUA_MODEL_ARTIFACT_CONFIG_MODEL_NAME = "_name_or_path"
 AQUA_MODEL_ARTIFACT_CONFIG_MODEL_TYPE = "model_type"
 AQUA_MODEL_ARTIFACT_FILE = "model_file"
+HF_METADATA_FOLDER = ".cache/"
 HF_LOGIN_DEFAULT_TIMEOUT = 2
 
 TRAINING_METRICS_FINAL = "training_metrics_final"

ads/aqua/evaluation/entities.py

@@ -64,6 +64,10 @@ class CreateAquaEvaluationDetails(Serializable):
         The metrics for the evaluation.
     force_overwrite: (bool, optional). Defaults to `False`.
         Whether to force overwrite the existing file in object storage.
+    freeform_tags: (dict, optional)
+        Freeform tags for the evaluation model
+    defined_tags: (dict, optional)
+        Defined tags for the evaluation model
     """
 
     evaluation_source_id: str
@@ -83,8 +87,10 @@ class CreateAquaEvaluationDetails(Serializable):
     ocpus: Optional[float] = None
     log_group_id: Optional[str] = None
     log_id: Optional[str] = None
-    metrics: Optional[List[str]] = None
+    metrics: Optional[List[Dict[str, Any]]] = None
     force_overwrite: Optional[bool] = False
+    freeform_tags: Optional[dict] = None
+    defined_tags: Optional[dict] = None
 
     class Config:
         extra = "ignore"
@@ -140,7 +146,7 @@ class AquaEvaluationCommands(Serializable):
     evaluation_id: str
     evaluation_target_id: str
     input_data: Dict[str, Any]
-    metrics: List[str]
+    metrics: List[Dict[str, Any]]
     output_dir: str
     params: Dict[str, Any]
 

ads/aqua/evaluation/evaluation.py

@@ -159,7 +159,8 @@ def create(
                 create_aqua_evaluation_details = CreateAquaEvaluationDetails(**kwargs)
             except Exception as ex:
                 custom_errors = {
-                    ".".join(map(str, e["loc"])): e["msg"] for e in json.loads(ex.json())
+                    ".".join(map(str, e["loc"])): e["msg"]
+                    for e in json.loads(ex.json())
                 }
                 raise AquaValueError(
                     f"Invalid create evaluation parameters. Error details: {custom_errors}."
@@ -296,6 +297,10 @@ def create(
                 evaluation_mvs_freeform_tags = {
                     Tags.AQUA_EVALUATION: Tags.AQUA_EVALUATION,
                 }
+                evaluation_mvs_freeform_tags = {
+                    **evaluation_mvs_freeform_tags,
+                    **(create_aqua_evaluation_details.freeform_tags or {}),
+                }
 
                 model_version_set = (
                     ModelVersionSet()
@@ -306,6 +311,9 @@ def create(
                         create_aqua_evaluation_details.experiment_description
                     )
                     .with_freeform_tags(**evaluation_mvs_freeform_tags)
+                    .with_defined_tags(
+                        **(create_aqua_evaluation_details.defined_tags or {})
+                    )
                     # TODO: decide what parameters will be needed
                     .create(**kwargs)
                 )
@@ -368,6 +376,10 @@ def create(
             Tags.AQUA_EVALUATION: Tags.AQUA_EVALUATION,
             Tags.AQUA_EVALUATION_MODEL_ID: evaluation_model.id,
         }
+        evaluation_job_freeform_tags = {
+            **evaluation_job_freeform_tags,
+            **(create_aqua_evaluation_details.freeform_tags or {}),
+        }
 
         evaluation_job = Job(name=evaluation_model.display_name).with_infrastructure(
             DataScienceJob()
@@ -378,6 +390,7 @@ def create(
             .with_shape_name(create_aqua_evaluation_details.shape_name)
             .with_block_storage_size(create_aqua_evaluation_details.block_storage_size)
             .with_freeform_tag(**evaluation_job_freeform_tags)
+            .with_defined_tag(**(create_aqua_evaluation_details.defined_tags or {}))
         )
         if (
             create_aqua_evaluation_details.memory_in_gbs
@@ -424,6 +437,7 @@ def create(
         evaluation_job_run = evaluation_job.run(
             name=evaluation_model.display_name,
             freeform_tags=evaluation_job_freeform_tags,
+            defined_tags=(create_aqua_evaluation_details.defined_tags or {}),
             wait=False,
         )
         logger.debug(
@@ -443,13 +457,20 @@ def create(
             for metadata in evaluation_model_custom_metadata.to_dict()["data"]
         ]
 
+        evaluation_model_freeform_tags = {
+            Tags.AQUA_EVALUATION: Tags.AQUA_EVALUATION,
+            **(create_aqua_evaluation_details.freeform_tags or {}),
+        }
+        evaluation_model_defined_tags = (
+            create_aqua_evaluation_details.defined_tags or {}
+        )
+
         self.ds_client.update_model(
             model_id=evaluation_model.id,
             update_model_details=UpdateModelDetails(
                 custom_metadata_list=updated_custom_metadata_list,
-                freeform_tags={
-                    Tags.AQUA_EVALUATION: Tags.AQUA_EVALUATION,
-                },
+                freeform_tags=evaluation_model_freeform_tags,
+                defined_tags=evaluation_model_defined_tags,
             ),
         )
 
@@ -523,6 +544,8 @@ def create(
                 "evaluation_job_id": evaluation_job.id,
                 "evaluation_source": create_aqua_evaluation_details.evaluation_source_id,
                 "evaluation_experiment_id": experiment_model_version_set_id,
+                **evaluation_model_freeform_tags,
+                **evaluation_model_defined_tags,
             },
             parameters=AquaEvalParams(),
         )
@@ -619,11 +642,6 @@ def _build_launch_cmd(
             evaluation_id=evaluation_id,
             evaluation_target_id=evaluation_source_id,
             input_data={
-                "columns": {
-                    "prompt": "prompt",
-                    "completion": "completion",
-                    "category": "category",
-                },
                 "format": Path(dataset_path).suffix,
                 "url": dataset_path,
             },

ads/aqua/extension/deployment_handler.py

@@ -59,7 +59,7 @@ def delete(self, model_deployment_id):
         return self.finish(AquaDeploymentApp().delete(model_deployment_id))
 
     @handle_exceptions
-    def put(self, *args, **kwargs):
+    def put(self, *args, **kwargs):  # noqa: ARG002
         """
         Handles put request for the activating and deactivating OCI datascience model deployments
         Raises
@@ -82,7 +82,7 @@ def put(self, *args, **kwargs):
             raise HTTPError(400, f"The request {self.request.path} is invalid.")
 
     @handle_exceptions
-    def post(self, *args, **kwargs):
+    def post(self, *args, **kwargs):  # noqa: ARG002
         """
         Handles post request for the deployment APIs
         Raises
@@ -132,6 +132,8 @@ def post(self, *args, **kwargs):
         private_endpoint_id = input_data.get("private_endpoint_id")
         container_image_uri = input_data.get("container_image_uri")
         cmd_var = input_data.get("cmd_var")
+        freeform_tags = input_data.get("freeform_tags")
+        defined_tags = input_data.get("defined_tags")
 
         self.finish(
             AquaDeploymentApp().create(
@@ -157,6 +159,8 @@ def post(self, *args, **kwargs):
                 private_endpoint_id=private_endpoint_id,
                 container_image_uri=container_image_uri,
                 cmd_var=cmd_var,
+                freeform_tags=freeform_tags,
+                defined_tags=defined_tags,
             )
         )
 
@@ -196,7 +200,7 @@ def validate_predict_url(endpoint):
             return False
 
     @handle_exceptions
-    def post(self, *args, **kwargs):
+    def post(self, *args, **kwargs):  # noqa: ARG002
         """
         Handles inference request for the Active Model Deployments
         Raises
@@ -262,7 +266,7 @@ def get(self, model_id):
         )
 
     @handle_exceptions
-    def post(self, *args, **kwargs):
+    def post(self, *args, **kwargs):  # noqa: ARG002
         """Handles post request for the deployment param handler API.
 
         Raises

ads/aqua/extension/model_handler.py

@@ -96,7 +96,7 @@ def list(self):
         )
 
     @handle_exceptions
-    def post(self, *args, **kwargs):
+    def post(self, *args, **kwargs):  # noqa: ARG002
         """
         Handles post request for the registering any Aqua model.
         Raises
@@ -129,6 +129,10 @@ def post(self, *args, **kwargs):
             str(input_data.get("download_from_hf", "false")).lower() == "true"
         )
         inference_container_uri = input_data.get("inference_container_uri")
+        allow_patterns = input_data.get("allow_patterns")
+        ignore_patterns = input_data.get("ignore_patterns")
+        freeform_tags = input_data.get("freeform_tags")
+        defined_tags = input_data.get("defined_tags")
 
         return self.finish(
             AquaModelApp().register(
@@ -141,6 +145,10 @@ def post(self, *args, **kwargs):
                 project_id=project_id,
                 model_file=model_file,
                 inference_container_uri=inference_container_uri,
+                allow_patterns=allow_patterns,
+                ignore_patterns=ignore_patterns,
+                freeform_tags=freeform_tags,
+                defined_tags=defined_tags,
             )
         )
 
@@ -166,11 +174,9 @@ def put(self, id):
 
         enable_finetuning = input_data.get("enable_finetuning")
         task = input_data.get("task")
-        app=AquaModelApp()
+        app = AquaModelApp()
         self.finish(
-            app.edit_registered_model(
-                id, inference_container, enable_finetuning, task
-            )
+            app.edit_registered_model(id, inference_container, enable_finetuning, task)
         )
         app.clear_model_details_cache(model_id=id)
 
@@ -214,7 +220,7 @@ def _find_matching_aqua_model(model_id: str) -> Optional[AquaModelSummary]:
         return None
 
     @handle_exceptions
-    def get(self, *args, **kwargs):
+    def get(self, *args, **kwargs):  # noqa: ARG002
         """
         Finds a list of matching models from hugging face based on query string provided from users.
 
@@ -235,7 +241,7 @@ def get(self, *args, **kwargs):
         return self.finish({"models": models})
 
     @handle_exceptions
-    def post(self, *args, **kwargs):
+    def post(self, *args, **kwargs):  # noqa: ARG002
         """Handles post request for the HF Models APIs
 
         Raises

ads/aqua/extension/ui_handler.py

@@ -68,6 +68,8 @@ def get(self, id=""):
             return self.list_buckets()
         elif paths.startswith("aqua/job/shapes"):
             return self.list_job_shapes()
+        elif paths.startswith("aqua/modeldeployment/shapes"):
+            return self.list_model_deployment_shapes()
         elif paths.startswith("aqua/vcn"):
             return self.list_vcn()
         elif paths.startswith("aqua/subnets"):
@@ -160,6 +162,15 @@ def list_job_shapes(self, **kwargs):
             AquaUIApp().list_job_shapes(compartment_id=compartment_id, **kwargs)
         )
 
+    def list_model_deployment_shapes(self, **kwargs):
+        """Lists model deployment shapes available in the specified compartment."""
+        compartment_id = self.get_argument("compartment_id", default=COMPARTMENT_OCID)
+        return self.finish(
+            AquaUIApp().list_model_deployment_shapes(
+                compartment_id=compartment_id, **kwargs
+            )
+        )
+
     def list_vcn(self, **kwargs):
         """Lists the virtual cloud networks (VCNs) in the specified compartment."""
         compartment_id = self.get_argument("compartment_id", default=COMPARTMENT_OCID)
@@ -255,8 +266,9 @@ def post(self, *args, **kwargs):
 __handlers__ = [
     ("logging/?([^/]*)", AquaUIHandler),
     ("compartments/?([^/]*)", AquaUIHandler),
-    # TODO: change url to evaluation/experiements/?([^/]*)
+    # TODO: change url to evaluation/experiments/?([^/]*)
     ("experiment/?([^/]*)", AquaUIHandler),
+    ("modeldeployment/?([^/]*)", AquaUIHandler),
     ("versionsets/?([^/]*)", AquaUIHandler),
     ("buckets/?([^/]*)", AquaUIHandler),
     ("job/shapes/?([^/]*)", AquaUIHandler),