Resolve merge conflicts

Author: VipulMascarenhas

.github/workflows/run-forecast-unit-tests.yml

@@ -56,6 +56,6 @@ jobs:
           $CONDA/bin/conda init
           source /home/runner/.bashrc
           pip install -r test-requirements-operators.txt
-          pip install "oracle-automlx[forecasting]>=24.4.0"
+          pip install "oracle-automlx[forecasting]>=24.4.1"
           pip install pandas>=2.2.0
           python -m pytest -v -p no:warnings --durations=5 tests/operators/forecast

ads/aqua/__init__.py

@@ -1,12 +1,12 @@
 #!/usr/bin/env python
-# -*- coding: utf-8 -*-
-# Copyright (c) 2024 Oracle and/or its affiliates.
+# Copyright (c) 2024, 2025 Oracle and/or its affiliates.
 # Licensed under the Universal Permissive License v 1.0 as shown at https://oss.oracle.com/licenses/upl/
 
 
 import os
+from logging import getLogger
 
-from ads import logger, set_auth
+from ads import set_auth
 from ads.aqua.common.utils import fetch_service_compartment
 from ads.config import OCI_RESOURCE_PRINCIPAL_VERSION
 
@@ -19,6 +19,7 @@ def get_logger_level():
     return level
 
 
+logger = getLogger(__name__)
 logger.setLevel(get_logger_level())
 
 
@@ -27,7 +28,6 @@ def set_log_level(log_level: str):
 
     log_level = log_level.upper()
     logger.setLevel(log_level.upper())
-    logger.handlers[0].setLevel(log_level)
 
 
 if OCI_RESOURCE_PRINCIPAL_VERSION:

ads/aqua/common/enums.py

@@ -52,6 +52,9 @@ class InferenceContainerTypeFamily(str, metaclass=ExtendedEnumMeta):
     AQUA_VLLM_CONTAINER_FAMILY = "odsc-vllm-serving"
     AQUA_TGI_CONTAINER_FAMILY = "odsc-tgi-serving"
     AQUA_LLAMA_CPP_CONTAINER_FAMILY = "odsc-llama-cpp-serving"
+
+
+class CustomInferenceContainerTypeFamily(str, metaclass=ExtendedEnumMeta):
     AQUA_TEI_CONTAINER_FAMILY = "odsc-tei-serving"
 
 

ads/aqua/common/utils.py

@@ -11,6 +11,7 @@
 import random
 import re
 import shlex
+import shutil
 import subprocess
 from datetime import datetime, timedelta
 from functools import wraps
@@ -21,6 +22,8 @@
 import fsspec
 import oci
 from cachetools import TTLCache, cached
+from huggingface_hub.constants import HF_HUB_CACHE
+from huggingface_hub.file_download import repo_folder_name
 from huggingface_hub.hf_api import HfApi, ModelInfo
 from huggingface_hub.utils import (
     GatedRepoError,
@@ -30,6 +33,7 @@
 )
 from oci.data_science.models import JobRun, Model
 from oci.object_storage.models import ObjectSummary
+from pydantic import ValidationError
 
 from ads.aqua.common.enums import (
     InferenceContainerParamType,
@@ -820,6 +824,48 @@ def upload_folder(
     return f"oci://{os_details.bucket}@{os_details.namespace}" + "/" + object_path
 
 
+def cleanup_local_hf_model_artifact(
+    model_name: str,
+    local_dir: str = None,
+):
+    """
+    Helper function that deletes local artifacts downloaded from Hugging Face to free up disk space.
+    Parameters
+    ----------
+    model_name (str): Name of the huggingface model
+    local_dir (str): Local directory where the object is downloaded
+
+    """
+    if local_dir and os.path.exists(local_dir):
+        model_dir = os.path.join(local_dir, model_name)
+        model_dir = (
+            os.path.dirname(model_dir)
+            if "/" in model_name or os.sep in model_name
+            else model_dir
+        )
+        shutil.rmtree(model_dir, ignore_errors=True)
+        if os.path.exists(model_dir):
+            logger.debug(
+                f"Could not delete local model artifact directory: {model_dir}"
+            )
+        else:
+            logger.debug(f"Deleted local model artifact directory: {model_dir}.")
+
+    hf_local_path = os.path.join(
+        HF_HUB_CACHE, repo_folder_name(repo_id=model_name, repo_type="model")
+    )
+    shutil.rmtree(hf_local_path, ignore_errors=True)
+
+    if os.path.exists(hf_local_path):
+        logger.debug(
+            f"Could not clear the local Hugging Face cache directory {hf_local_path} for the model {model_name}."
+        )
+    else:
+        logger.debug(
+            f"Cleared contents of local Hugging Face cache directory {hf_local_path} for the model {model_name}."
+        )
+
+
 def is_service_managed_container(container):
     return container and container.startswith(SERVICE_MANAGED_CONTAINER_URI_SCHEME)
 
@@ -1161,3 +1207,15 @@ def validate_cmd_var(cmd_var: List[str], overrides: List[str]) -> List[str]:
 
     combined_cmd_var = cmd_var + overrides
     return combined_cmd_var
+
+
+def build_pydantic_error_message(ex: ValidationError):
+    """Added to handle error messages from pydantic model validator.
+    Combine both loc and msg for errors where loc (field) is present in error details, else only build error
+    message using msg field."""
+
+    return {
+        ".".join(map(str, e["loc"])): e["msg"]
+        for e in ex.errors()
+        if "loc" in e and e["loc"]
+    } or "; ".join(e["msg"] for e in ex.errors())

ads/aqua/data.py

@@ -1,9 +1,8 @@
 #!/usr/bin/env python
-# -*- coding: utf-8 -*-
-# Copyright (c) 2024 Oracle and/or its affiliates.
+# Copyright (c) 2024, 2025 Oracle and/or its affiliates.
 # Licensed under the Universal Permissive License v 1.0 as shown at https://oss.oracle.com/licenses/upl/
 
-from dataclasses import dataclass, field
+from dataclasses import dataclass
 
 from ads.common.serializer import DataClassSerializable
 
@@ -13,19 +12,3 @@ class AquaResourceIdentifier(DataClassSerializable):
     id: str = ""
     name: str = ""
     url: str = ""
-
-
-@dataclass(repr=False)
-class AquaJobSummary(DataClassSerializable):
-    """Represents an Aqua job summary."""
-
-    id: str
-    name: str
-    console_url: str
-    lifecycle_state: str
-    lifecycle_details: str
-    time_created: str
-    tags: dict
-    experiment: AquaResourceIdentifier = field(default_factory=AquaResourceIdentifier)
-    source: AquaResourceIdentifier = field(default_factory=AquaResourceIdentifier)
-    job: AquaResourceIdentifier = field(default_factory=AquaResourceIdentifier)

ads/aqua/extension/finetune_handler.py

@@ -10,9 +10,7 @@
 from ads.aqua.common.decorator import handle_exceptions
 from ads.aqua.extension.base_handler import AquaAPIhandler
 from ads.aqua.extension.errors import Errors
-from ads.aqua.extension.utils import validate_function_parameters
 from ads.aqua.finetuning import AquaFineTuningApp
-from ads.aqua.finetuning.entities import CreateFineTuningDetails
 
 
 class AquaFineTuneHandler(AquaAPIhandler):
@@ -49,11 +47,7 @@ def post(self, *args, **kwargs):  # noqa: ARG002
         if not input_data:
             raise HTTPError(400, Errors.NO_INPUT_DATA)
 
-        validate_function_parameters(
-            data_class=CreateFineTuningDetails, input_data=input_data
-        )
-
-        self.finish(AquaFineTuningApp().create(CreateFineTuningDetails(**input_data)))
+        self.finish(AquaFineTuningApp().create(**input_data))
 
     def get_finetuning_config(self, model_id):
         """Gets the finetuning config for Aqua model."""

ads/aqua/extension/model_handler.py

@@ -1,5 +1,5 @@
 #!/usr/bin/env python
-# Copyright (c) 2024 Oracle and/or its affiliates.
+# Copyright (c) 2024, 2025 Oracle and/or its affiliates.
 # Licensed under the Universal Permissive License v 1.0 as shown at https://oss.oracle.com/licenses/upl/
 
 from typing import Optional
@@ -8,6 +8,9 @@
 from tornado.web import HTTPError
 
 from ads.aqua.common.decorator import handle_exceptions
+from ads.aqua.common.enums import (
+    CustomInferenceContainerTypeFamily,
+)
 from ads.aqua.common.errors import AquaRuntimeError, AquaValueError
 from ads.aqua.common.utils import (
     get_hf_model_info,
@@ -128,6 +131,10 @@ def post(self, *args, **kwargs):  # noqa: ARG002
         download_from_hf = (
             str(input_data.get("download_from_hf", "false")).lower() == "true"
         )
+        local_dir = input_data.get("local_dir")
+        cleanup_model_cache = (
+            str(input_data.get("cleanup_model_cache", "true")).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")
@@ -143,6 +150,8 @@ def post(self, *args, **kwargs):  # noqa: ARG002
                 model=model,
                 os_path=os_path,
                 download_from_hf=download_from_hf,
+                local_dir=local_dir,
+                cleanup_model_cache=cleanup_model_cache,
                 inference_container=inference_container,
                 finetuning_container=finetuning_container,
                 compartment_id=compartment_id,
@@ -168,7 +177,9 @@ def put(self, id):
             raise HTTPError(400, Errors.NO_INPUT_DATA)
 
         inference_container = input_data.get("inference_container")
+        inference_container_uri = input_data.get("inference_container_uri")
         inference_containers = AquaModelApp.list_valid_inference_containers()
+        inference_containers.extend(CustomInferenceContainerTypeFamily.values())
         if (
             inference_container is not None
             and inference_container not in inference_containers
@@ -181,7 +192,13 @@ def put(self, id):
         task = input_data.get("task")
         app = AquaModelApp()
         self.finish(
-            app.edit_registered_model(id, inference_container, enable_finetuning, task)
+            app.edit_registered_model(
+                id,
+                inference_container,
+                inference_container_uri,
+                enable_finetuning,
+                task,
+            )
         )
         app.clear_model_details_cache(model_id=id)
 

ads/aqua/finetuning/constants.py

@@ -1,6 +1,5 @@
 #!/usr/bin/env python
-# -*- coding: utf-8 -*-
-# Copyright (c) 2024 Oracle and/or its affiliates.
+# Copyright (c) 2024, 2025 Oracle and/or its affiliates.
 # Licensed under the Universal Permissive License v 1.0 as shown at https://oss.oracle.com/licenses/upl/
 
 from ads.common.extended_enum import ExtendedEnumMeta
@@ -17,4 +16,8 @@ class FineTuneCustomMetadata(str, metaclass=ExtendedEnumMeta):
     SERVICE_MODEL_FINE_TUNE_CONTAINER = "finetune-container"
 
 
+class FineTuningRestrictedParams(str, metaclass=ExtendedEnumMeta):
+    OPTIMIZER = "optimizer"
+
+
 ENV_AQUA_FINE_TUNING_CONTAINER = "AQUA_FINE_TUNING_CONTAINER"

ads/aqua/finetuning/entities.py

@@ -1,18 +1,24 @@
 #!/usr/bin/env python
-# Copyright (c) 2024 Oracle and/or its affiliates.
+# Copyright (c) 2024, 2025 Oracle and/or its affiliates.
 # Licensed under the Universal Permissive License v 1.0 as shown at https://oss.oracle.com/licenses/upl/
-from dataclasses import dataclass, field
-from typing import List, Optional
 
-from ads.aqua.data import AquaJobSummary
-from ads.common.serializer import DataClassSerializable
+import json
+from typing import List, Literal, Optional, Union
 
+from pydantic import Field, model_validator
 
-@dataclass(repr=False)
-class AquaFineTuningParams(DataClassSerializable):
-    epochs: int
+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 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
     )
@@ -22,21 +28,59 @@ class AquaFineTuningParams(DataClassSerializable):
     lora_alpha: Optional[int] = None
     lora_dropout: Optional[float] = None
     lora_target_linear: Optional[bool] = None
-    lora_target_modules: Optional[List] = None
+    lora_target_modules: Optional[List[str]] = None
     early_stopping_patience: Optional[int] = None
     early_stopping_threshold: Optional[float] = None
 
+    class Config:
+        extra = "allow"
+
+    def to_dict(self) -> dict:
+        return json.loads(super().to_json(exclude_none=True))
+
+    @model_validator(mode="before")
+    @classmethod
+    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
+        restricted_params = [
+            param for param in data if param in FineTuningRestrictedParams.values()
+        ]
+        if restricted_params:
+            raise AquaValueError(
+                f"Found restricted parameter name: {restricted_params}"
+            )
+        return data
 
-@dataclass(repr=False)
-class AquaFineTuningSummary(AquaJobSummary, DataClassSerializable):
-    parameters: AquaFineTuningParams = field(default_factory=AquaFineTuningParams)
 
+class AquaFineTuningSummary(Serializable):
+    """Represents a summary of Aqua Finetuning job."""
 
-@dataclass(repr=False)
-class CreateFineTuningDetails(DataClassSerializable):
-    """Dataclass to create aqua model fine tuning.
+    id: str
+    name: str
+    console_url: str
+    lifecycle_state: str
+    lifecycle_details: str
+    time_created: str
+    tags: dict
+    experiment: AquaResourceIdentifier = Field(default_factory=AquaResourceIdentifier)
+    source: AquaResourceIdentifier = Field(default_factory=AquaResourceIdentifier)
+    job: AquaResourceIdentifier = Field(default_factory=AquaResourceIdentifier)
+    parameters: AquaFineTuningParams = Field(default_factory=AquaFineTuningParams)
 
-    Fields
+    class Config:
+        extra = "ignore"
+
+    def to_dict(self) -> dict:
+        return json.loads(super().to_json(exclude_none=True))
+
+
+class CreateFineTuningDetails(Serializable):
+    """Class to create aqua model fine-tuning instance.
+
+    Properties
     ------
     ft_source_id: str
         The fine tuning source id. Must be model ocid.
@@ -107,3 +151,6 @@ class CreateFineTuningDetails(DataClassSerializable):
     force_overwrite: Optional[bool] = False
     freeform_tags: Optional[dict] = None
     defined_tags: Optional[dict] = None
+
+    class Config:
+        extra = "ignore"