Merge remote-tracking branch 'origin/feature/autogen' into feature/au…

…togen

ads/aqua/common/decorator.py

@@ -69,6 +69,16 @@ def inner_function(
                 reason=error.message,
                 service_payload=error.args[0] if error.args else None,
                 exc_info=sys.exc_info(),
+                aqua_api_details=dict(
+                    # __qualname__ gives information of class and name of api
+                    aqua_api_name=func.__qualname__,
+                    oci_api_name=getattr(
+                        error, "operation_name", "Unknown OCI Operation"
+                    ),
+                    service_endpoint=getattr(
+                        error, "request_endpoint", "Unknown Request Endpoint"
+                    )
+                )
             )
         except (
             ClientError,

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

@@ -9,11 +9,12 @@
 This module contains dataclasses for aqua evaluation.
 """
 
+from typing import Any, Dict, List, Optional
+
 from pydantic import Field
-from typing import Any, Dict, List, Optional, Union
 
-from ads.aqua.data import AquaResourceIdentifier
 from ads.aqua.config.utils.serializer import Serializable
+from ads.aqua.data import AquaResourceIdentifier
 
 
 class CreateAquaEvaluationDetails(Serializable):
@@ -87,6 +88,8 @@ class CreateAquaEvaluationDetails(Serializable):
 
     class Config:
         extra = "ignore"
+        protected_namespaces = ()
+
 
 class AquaEvalReport(Serializable):
     evaluation_id: str = ""
@@ -95,6 +98,7 @@ class AquaEvalReport(Serializable):
     class Config:
         extra = "ignore"
 
+
 class AquaEvalParams(Serializable):
     shape: str = ""
     dataset_path: str = ""
@@ -103,6 +107,7 @@ class AquaEvalParams(Serializable):
     class Config:
         extra = "allow"
 
+
 class AquaEvalMetric(Serializable):
     key: str
     name: str
@@ -111,6 +116,7 @@ class AquaEvalMetric(Serializable):
     class Config:
         extra = "ignore"
 
+
 class AquaEvalMetricSummary(Serializable):
     metric: str = ""
     score: str = ""
@@ -119,6 +125,7 @@ class AquaEvalMetricSummary(Serializable):
     class Config:
         extra = "ignore"
 
+
 class AquaEvalMetrics(Serializable):
     id: str
     report: str
@@ -128,6 +135,7 @@ class AquaEvalMetrics(Serializable):
     class Config:
         extra = "ignore"
 
+
 class AquaEvaluationCommands(Serializable):
     evaluation_id: str
     evaluation_target_id: str
@@ -139,6 +147,7 @@ class AquaEvaluationCommands(Serializable):
     class Config:
         extra = "ignore"
 
+
 class AquaEvaluationSummary(Serializable):
     """Represents a summary of Aqua evalution."""
 
@@ -157,6 +166,7 @@ class AquaEvaluationSummary(Serializable):
     class Config:
         extra = "ignore"
 
+
 class AquaEvaluationDetail(AquaEvaluationSummary):
     """Represents a details of Aqua evalution."""
 

ads/aqua/extension/aqua_ws_msg_handler.py

@@ -78,10 +78,12 @@ def write_error(self, status_code, **kwargs):
                 logger.warning(reply["message"])
         # telemetry may not be present if there is an error while initializing
         if hasattr(self, "telemetry"):
+            aqua_api_details = kwargs.get("aqua_api_details", {})
             self.telemetry.record_event_async(
                 category="aqua/error",
                 action=str(status_code),
                 value=reason,
+                **aqua_api_details
             )
         response = AquaWsError(
             status=status_code,

ads/aqua/extension/base_handler.py

@@ -98,10 +98,12 @@ def write_error(self, status_code, **kwargs):
 
         # telemetry may not be present if there is an error while initializing
         if hasattr(self, "telemetry"):
+            aqua_api_details = kwargs.get("aqua_api_details", {})
             self.telemetry.record_event_async(
                 category="aqua/error",
                 action=str(status_code),
                 value=reason,
+                **aqua_api_details
             )
 
         self.finish(json.dumps(reply))

ads/aqua/extension/model_handler.py

@@ -129,6 +129,8 @@ 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")
 
         return self.finish(
             AquaModelApp().register(
@@ -141,6 +143,8 @@ 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,
             )
         )
 

ads/aqua/finetuning/constants.py

@@ -15,3 +15,6 @@ class FineTuneCustomMetadata(str, metaclass=ExtendedEnumMeta):
     SERVICE_MODEL_ARTIFACT_LOCATION = "artifact_location"
     SERVICE_MODEL_DEPLOYMENT_CONTAINER = "deployment-container"
     SERVICE_MODEL_FINE_TUNE_CONTAINER = "finetune-container"
+
+
+ENV_AQUA_FINE_TUNING_CONTAINER = "AQUA_FINE_TUNING_CONTAINER"

ads/aqua/finetuning/finetuning.py

@@ -31,7 +31,10 @@
     UNKNOWN_DICT,
 )
 from ads.aqua.data import AquaResourceIdentifier
-from ads.aqua.finetuning.constants import *
+from ads.aqua.finetuning.constants import (
+    ENV_AQUA_FINE_TUNING_CONTAINER,
+    FineTuneCustomMetadata,
+)
 from ads.aqua.finetuning.entities import *
 from ads.common.auth import default_signer
 from ads.common.object_storage_details import ObjectStorageDetails
@@ -310,6 +313,15 @@ def create(
         except Exception:
             pass
 
+        if not is_custom_container and ENV_AQUA_FINE_TUNING_CONTAINER in os.environ:
+            ft_container = os.environ[ENV_AQUA_FINE_TUNING_CONTAINER]
+            logger.info(
+                "Using container set by environment variable %s=%s",
+                ENV_AQUA_FINE_TUNING_CONTAINER,
+                ft_container,
+            )
+            is_custom_container = True
+
         ft_parameters.batch_size = ft_parameters.batch_size or (
             ft_config.get("shape", UNKNOWN_DICT)
             .get(create_fine_tuning_details.shape_name, UNKNOWN_DICT)
@@ -559,7 +571,6 @@ def get_finetuning_config(self, model_id: str) -> Dict:
         Dict:
             A dict of allowed finetuning configs.
         """
-
         config = self.get_config(model_id, AQUA_MODEL_FINETUNING_CONFIG)
         if not config:
             logger.debug(

ads/aqua/model/entities.py

@@ -289,6 +289,8 @@ class ImportModelDetails(CLIBuilderMixin):
     project_id: Optional[str] = None
     model_file: Optional[str] = None
     inference_container_uri: Optional[str] = None
+    allow_patterns: Optional[List[str]] = None
+    ignore_patterns: Optional[List[str]] = None
 
     def __post_init__(self):
         self._command = "model register"

ads/aqua/model/model.py

@@ -40,6 +40,7 @@
     AQUA_MODEL_ARTIFACT_CONFIG_MODEL_TYPE,
     AQUA_MODEL_ARTIFACT_FILE,
     AQUA_MODEL_TYPE_CUSTOM,
+    HF_METADATA_FOLDER,
     LICENSE_TXT,
     MODEL_BY_REFERENCE_OSS_PATH_KEY,
     README,
@@ -1274,6 +1275,8 @@ def _download_model_from_hf(
         model_name: str,
         os_path: str,
         local_dir: str = None,
+        allow_patterns: List[str] = None,
+        ignore_patterns: List[str] = None,
     ) -> str:
         """This helper function downloads the model artifact from Hugging Face to a local folder, then uploads
         to object storage location.
@@ -1283,6 +1286,12 @@ def _download_model_from_hf(
         model_name (str): The huggingface model name.
         os_path (str): The OS path where the model files are located.
         local_dir (str): The local temp dir to store the huggingface model.
+        allow_patterns (list): Model files matching at least one pattern are downloaded.
+            Example: ["*.json"] will download all .json files. ["folder/*"] will download all files under `folder`.
+            Patterns are Standard Wildcards (globbing patterns) and rules can be found here: https://docs.python.org/3/library/fnmatch.html
+        ignore_patterns (list): Model files matching any of the patterns are not downloaded.
+            Example: ["*.json"] will ignore all .json files. ["folder/*"] will ignore all files under `folder`.
+            Patterns are Standard Wildcards (globbing patterns) and rules can be found here: https://docs.python.org/3/library/fnmatch.html
 
         Returns
         -------
@@ -1293,30 +1302,19 @@ def _download_model_from_hf(
         if not local_dir:
             local_dir = os.path.join(os.path.expanduser("~"), "cached-model")
         local_dir = os.path.join(local_dir, model_name)
-        retry = 10
-        i = 0
-        huggingface_download_err_message = None
-        while i < retry:
-            try:
-                # Download to cache folder. The while loop retries when there is a network failure
-                snapshot_download(repo_id=model_name)
-            except Exception as e:
-                huggingface_download_err_message = str(e)
-                i += 1
-            else:
-                break
-        if i == retry:
-            raise Exception(
-                f"Could not download the model {model_name} from https://huggingface.co with message {huggingface_download_err_message}"
-            )
         os.makedirs(local_dir, exist_ok=True)
-        # Copy the model from the cache to destination
-        snapshot_download(repo_id=model_name, local_dir=local_dir)
-        # Upload to object storage
+        snapshot_download(
+            repo_id=model_name,
+            local_dir=local_dir,
+            allow_patterns=allow_patterns,
+            ignore_patterns=ignore_patterns,
+        )
+        # Upload to object storage and skip .cache/huggingface/ folder
         model_artifact_path = upload_folder(
             os_path=os_path,
             local_dir=local_dir,
             model_name=model_name,
+            exclude_pattern=f"{HF_METADATA_FOLDER}*"
         )
 
         return model_artifact_path
@@ -1335,6 +1333,12 @@ def register(
                 os_path (str): Object storage destination URI to store the downloaded model. Format: oci://bucket-name@namespace/prefix
                 inference_container (str): selects service defaults
                 finetuning_container (str): selects service defaults
+                allow_patterns (list): Model files matching at least one pattern are downloaded.
+                    Example: ["*.json"] will download all .json files. ["folder/*"] will download all files under `folder`.
+                    Patterns are Standard Wildcards (globbing patterns) and rules can be found here: https://docs.python.org/3/library/fnmatch.html
+                ignore_patterns (list): Model files matching any of the patterns are not downloaded.
+                    Example: ["*.json"] will ignore all .json files. ["folder/*"] will ignore all files under `folder`.
+                    Patterns are Standard Wildcards (globbing patterns) and rules can be found here: https://docs.python.org/3/library/fnmatch.html
 
         Returns:
             AquaModel:
@@ -1381,6 +1385,8 @@ def register(
                 model_name=model_name,
                 os_path=import_model_details.os_path,
                 local_dir=import_model_details.local_dir,
+                allow_patterns=import_model_details.allow_patterns,
+                ignore_patterns=import_model_details.ignore_patterns,
             ).rstrip("/")
         else:
             artifact_path = import_model_details.os_path.rstrip("/")

ads/opctl/operator/lowcode/anomaly/model/anomaly_merlion.py

@@ -4,9 +4,11 @@
 # Licensed under the Universal Permissive License v 1.0 as shown at https://oss.oracle.com/licenses/upl/
 
 import importlib
+import logging
 
 import numpy as np
 import pandas as pd
+import report_creator as rc
 from merlion.post_process.threshold import AggregateAlarms
 from merlion.utils import TimeSeries
 
@@ -21,6 +23,8 @@
 from .anomaly_dataset import AnomalyOutput
 from .base_model import AnomalyOperatorBaseModel
 
+logging.getLogger("report_creator").setLevel(logging.WARNING)
+
 
 class AnomalyMerlionOperatorModel(AnomalyOperatorBaseModel):
     """Class representing Merlion Anomaly Detection operator model."""
@@ -84,7 +88,7 @@ def _build_model(self) -> AnomalyOutput:
         for target, df in self.datasets.full_data_dict.items():
             data = df.set_index(date_column)
             data = TimeSeries.from_pd(data)
-            for model_name, (model_config, model) in model_config_map.items():
+            for _, (model_config, model) in model_config_map.items():
                 if self.spec.model == SupportedModels.BOCPD:
                     model_config = model_config(**self.spec.model_kwargs)
                 else:
@@ -115,7 +119,7 @@ def _build_model(self) -> AnomalyOutput:
                     y_pred = (y_pred.to_pd().reset_index()["anom_score"] > 0).astype(
                         int
                     )
-                except Exception as e:
+                except Exception:
                     y_pred = (
                         scores["anom_score"]
                         > np.percentile(
@@ -135,15 +139,12 @@ def _build_model(self) -> AnomalyOutput:
                         OutputColumns.SCORE_COL: scores["anom_score"],
                     }
                 ).reset_index(drop=True)
-                # model_objects[model_name].append(model)
 
                 anomaly_output.add_output(target, anomaly, score)
         return anomaly_output
 
     def _generate_report(self):
         """Genreates a report for the model."""
-        import report_creator as rc
-
         other_sections = [
             rc.Heading("Selected Models Overview", level=2),
             rc.Text(