oracle-ads 2.11.14__tar.gz → 2.11.16__tar.gz

{oracle_ads-2.11.14 → oracle_ads-2.11.16}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: oracle_ads
-Version: 2.11.14
+Version: 2.11.16
 Summary: Oracle Accelerated Data Science SDK
 Keywords: Oracle Cloud Infrastructure,OCI,Machine Learning,ML,Artificial Intelligence,AI,Data Science,Cloud,Oracle
 Author: Oracle Data Science
@@ -114,6 +114,10 @@ Requires-Dist: scrubadub_spacy ; extra == "pii"
 Requires-Dist: spacy-transformers==1.2.5 ; extra == "pii"
 Requires-Dist: spacy==3.6.1 ; extra == "pii"
 Requires-Dist: report-creator==1.0.9 ; extra == "pii"
+Requires-Dist: oracle_ads[opctl] ; extra == "recommender"
+Requires-Dist: scikit-surprise ; extra == "recommender"
+Requires-Dist: plotly ; extra == "recommender"
+Requires-Dist: report-creator==1.0.9 ; extra == "recommender"
 Requires-Dist: pyspark>=3.0.0 ; extra == "spark"
 Requires-Dist: oracle_ads[viz] ; extra == "tensorflow"
 Requires-Dist: tensorflow<=2.15.1 ; extra == "tensorflow"
@@ -163,6 +167,7 @@ Provides-Extra: onnx
 Provides-Extra: opctl
 Provides-Extra: optuna
 Provides-Extra: pii
+Provides-Extra: recommender
 Provides-Extra: spark
 Provides-Extra: tensorflow
 Provides-Extra: testsuite

oracle_ads-2.11.16/ads/aqua/common/entities.py

@@ -0,0 +1,17 @@
+#!/usr/bin/env python
+# 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/
+
+
+class ContainerSpec:
+    """
+    Class to hold to hold keys within the container spec.
+    """
+
+    CONTAINER_SPEC = "containerSpec"
+    CLI_PARM = "cliParam"
+    SERVER_PORT = "serverPort"
+    HEALTH_CHECK_PORT = "healthCheckPort"
+    ENV_VARS = "envVars"
+    RESTRICTED_PARAMS = "restrictedParams"
+    EVALUATION_CONFIGURATION = "evaluationConfiguration"

{oracle_ads-2.11.14 → oracle_ads-2.11.16}/ads/aqua/common/enums.py

@@ -1,5 +1,4 @@
 #!/usr/bin/env python
-# -*- coding: utf-8 -*-
 # 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/
 
@@ -8,6 +7,7 @@ aqua.common.enums
 ~~~~~~~~~~~~~~
 This module contains the set of enums used in AQUA.
 """
+
 from ads.common.extended_enum import ExtendedEnumMeta
 
 
@@ -38,21 +38,25 @@ class Tags(str, metaclass=ExtendedEnumMeta):
     READY_TO_IMPORT = "ready_to_import"
     BASE_MODEL_CUSTOM = "aqua_custom_base_model"
     AQUA_EVALUATION_MODEL_ID = "evaluation_model_id"
+    MODEL_FORMAT = "model_format"
 
 
 class InferenceContainerType(str, metaclass=ExtendedEnumMeta):
     CONTAINER_TYPE_VLLM = "vllm"
     CONTAINER_TYPE_TGI = "tgi"
+    CONTAINER_TYPE_LLAMA_CPP = "llama-cpp"
 
 
 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 InferenceContainerParamType(str, metaclass=ExtendedEnumMeta):
     PARAM_TYPE_VLLM = "VLLM_PARAMS"
     PARAM_TYPE_TGI = "TGI_PARAMS"
+    PARAM_TYPE_LLAMA_CPP = "LLAMA_CPP_PARAMS"
 
 
 class HuggingFaceTags(str, metaclass=ExtendedEnumMeta):

{oracle_ads-2.11.14 → oracle_ads-2.11.16}/ads/aqua/common/utils.py

@@ -1,8 +1,8 @@
 #!/usr/bin/env python
-# -*- coding: utf-8 -*-
 # 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/
 """AQUA utils and constants."""
+
 import asyncio
 import base64
 import json
@@ -10,6 +10,7 @@ import logging
 import os
 import random
 import re
+from datetime import datetime, timedelta
 from functools import wraps
 from pathlib import Path
 from string import Template
@@ -17,18 +18,36 @@ from typing import List, Union
 
 import fsspec
 import oci
+from cachetools import TTLCache, cached
 from oci.data_science.models import JobRun, Model
+from oci.object_storage.models import ObjectSummary
 
-from ads.aqua.common.enums import RqsAdditionalDetails
+from ads.aqua.common.enums import (
+    InferenceContainerParamType,
+    InferenceContainerType,
+    RqsAdditionalDetails,
+)
 from ads.aqua.common.errors import (
     AquaFileNotFoundError,
     AquaRuntimeError,
     AquaValueError,
 )
-from ads.aqua.constants import *
+from ads.aqua.constants import (
+    AQUA_GA_LIST,
+    COMPARTMENT_MAPPING_KEY,
+    CONSOLE_LINK_RESOURCE_TYPE_MAPPING,
+    CONTAINER_INDEX,
+    MAXIMUM_ALLOWED_DATASET_IN_BYTE,
+    MODEL_BY_REFERENCE_OSS_PATH_KEY,
+    SERVICE_MANAGED_CONTAINER_URI_SCHEME,
+    SUPPORTED_FILE_FORMATS,
+    TGI_INFERENCE_RESTRICTED_PARAMS,
+    UNKNOWN,
+    UNKNOWN_JSON_STR,
+    VLLM_INFERENCE_RESTRICTED_PARAMS,
+)
 from ads.aqua.data import AquaResourceIdentifier
 from ads.common.auth import default_signer
-from ads.common.decorator.threaded import threaded
 from ads.common.extended_enum import ExtendedEnumMeta
 from ads.common.object_storage_details import ObjectStorageDetails
 from ads.common.oci_resource import SEARCH_TYPE, OCIResource
@@ -74,15 +93,15 @@ class LifecycleStatus(str, metaclass=ExtendedEnumMeta):
 
         status = LifecycleStatus.UNKNOWN
         if evaluation_status == Model.LIFECYCLE_STATE_ACTIVE:
-            if (
-                job_run_status == JobRun.LIFECYCLE_STATE_IN_PROGRESS
-                or job_run_status == JobRun.LIFECYCLE_STATE_ACCEPTED
-            ):
+            if job_run_status in {
+                JobRun.LIFECYCLE_STATE_IN_PROGRESS,
+                JobRun.LIFECYCLE_STATE_ACCEPTED,
+            }:
                 status = JobRun.LIFECYCLE_STATE_IN_PROGRESS
-            elif (
-                job_run_status == JobRun.LIFECYCLE_STATE_FAILED
-                or job_run_status == JobRun.LIFECYCLE_STATE_NEEDS_ATTENTION
-            ):
+            elif job_run_status in {
+                JobRun.LIFECYCLE_STATE_FAILED,
+                JobRun.LIFECYCLE_STATE_NEEDS_ATTENTION,
+            }:
                 status = JobRun.LIFECYCLE_STATE_FAILED
             else:
                 status = job_run_status
@@ -196,13 +215,9 @@ def read_file(file_path: str, **kwargs) -> str:
         return UNKNOWN
 
 
-@threaded()
 def load_config(file_path: str, config_file_name: str, **kwargs) -> dict:
     artifact_path = f"{file_path.rstrip('/')}/{config_file_name}"
-    if artifact_path.startswith("oci://"):
-        signer = default_signer()
-    else:
-        signer = {}
+    signer = default_signer() if artifact_path.startswith("oci://") else {}
     config = json.loads(
         read_file(file_path=artifact_path, auth=signer, **kwargs) or UNKNOWN_JSON_STR
     )
@@ -214,6 +229,32 @@ def load_config(file_path: str, config_file_name: str, **kwargs) -> dict:
     return config
 
 
+def list_os_files_with_extension(oss_path: str, extension: str) -> [str]:
+    """
+    List files in the specified directory with the given extension.
+
+    Parameters:
+    - oss_path: The path to the directory where files are located.
+    - extension: The file extension to filter by (e.g., 'txt' for text files).
+
+    Returns:
+    - A list of file paths matching the specified extension.
+    """
+
+    oss_client = ObjectStorageDetails.from_path(oss_path)
+
+    # Ensure the extension is prefixed with a dot if not already
+    if not extension.startswith("."):
+        extension = "." + extension
+    files: List[ObjectSummary] = oss_client.list_objects().objects
+
+    return [
+        file.name[len(oss_client.filepath) :].lstrip("/")
+        for file in files
+        if file.name.endswith(extension)
+    ]
+
+
 def is_valid_ocid(ocid: str) -> bool:
     """Checks if the given ocid is valid.
 
@@ -448,7 +489,7 @@ def _build_resource_identifier(
 
 
 def _get_experiment_info(
-    model: Union[oci.resource_search.models.ResourceSummary, DataScienceModel]
+    model: Union[oci.resource_search.models.ResourceSummary, DataScienceModel],
 ) -> tuple:
     """Returns ocid and name of the experiment."""
     return (
@@ -489,6 +530,7 @@ def container_config_path():
     return f"oci://{AQUA_SERVICE_MODELS_BUCKET}@{CONDA_BUCKET_NS}/service_models/config"
 
 
+@cached(cache=TTLCache(maxsize=1, ttl=timedelta(hours=5), timer=datetime.now))
 def get_container_config():
     config = load_config(
         file_path=container_config_path(),
@@ -609,7 +651,7 @@ def extract_id_and_name_from_tag(tag: str):
     base_model_name = UNKNOWN
     try:
         base_model_ocid, base_model_name = tag.split("#")
-    except:
+    except Exception:
         pass
 
     if not (is_valid_ocid(base_model_ocid) and base_model_name):
@@ -646,7 +688,7 @@ def get_resource_name(ocid: str) -> str:
     try:
         resource = query_resource(ocid, return_all=False)
         name = resource.display_name if resource else UNKNOWN
-    except:
+    except Exception:
         name = UNKNOWN
     return name
 
@@ -670,8 +712,8 @@ def get_model_by_reference_paths(model_file_description: dict):
 
     if not models:
         raise AquaValueError(
-            f"Model path is not available in the model json artifact. "
-            f"Please check if the model created by reference has the correct artifact."
+            "Model path is not available in the model json artifact. "
+            "Please check if the model created by reference has the correct artifact."
         )
 
     if len(models) > 0:
@@ -848,3 +890,48 @@ def copy_model_config(artifact_path: str, os_path: str, auth: dict = None):
     except Exception as ex:
         logger.debug(ex)
         logger.debug(f"Failed to copy config folder from {artifact_path} to {os_path}.")
+
+
+def get_container_params_type(container_type_name: str) -> str:
+    """The utility function accepts the deployment container type name and returns the corresponding params name.
+    Parameters
+    ----------
+    container_type_name: str
+        type of deployment container, like odsc-vllm-serving or odsc-tgi-serving.
+
+    Returns
+    -------
+        InferenceContainerParamType value
+
+    """
+    # check substring instead of direct match in case container_type_name changes in the future
+    if InferenceContainerType.CONTAINER_TYPE_VLLM in container_type_name.lower():
+        return InferenceContainerParamType.PARAM_TYPE_VLLM
+    elif InferenceContainerType.CONTAINER_TYPE_TGI in container_type_name.lower():
+        return InferenceContainerParamType.PARAM_TYPE_TGI
+    elif InferenceContainerType.CONTAINER_TYPE_LLAMA_CPP in container_type_name.lower():
+        return InferenceContainerParamType.PARAM_TYPE_LLAMA_CPP
+    else:
+        return UNKNOWN
+
+
+def get_restricted_params_by_container(container_type_name: str) -> set:
+    """The utility function accepts the deployment container type name and returns a set of restricted params
+        for that container.
+    Parameters
+    ----------
+    container_type_name: str
+        type of deployment container, like odsc-vllm-serving or odsc-tgi-serving.
+
+    Returns
+    -------
+        Set of restricted params based on container type
+
+    """
+    # check substring instead of direct match in case container_type_name changes in the future
+    if InferenceContainerType.CONTAINER_TYPE_VLLM in container_type_name.lower():
+        return VLLM_INFERENCE_RESTRICTED_PARAMS
+    elif InferenceContainerType.CONTAINER_TYPE_TGI in container_type_name.lower():
+        return TGI_INFERENCE_RESTRICTED_PARAMS
+    else:
+        return set()

{oracle_ads-2.11.14 → oracle_ads-2.11.16}/ads/aqua/config/config.py

@@ -1,5 +1,4 @@
 #!/usr/bin/env python
-# -*- coding: utf-8 -*-
 # 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/
 
@@ -14,5 +13,6 @@ def get_finetuning_config_defaults():
             "BM.GPU.A10.4": {"batch_size": 1, "replica": 1},
             "BM.GPU4.8": {"batch_size": 4, "replica": 1},
             "BM.GPU.A100-v2.8": {"batch_size": 6, "replica": 1},
+            "BM.GPU.H100.8": {"batch_size": 6, "replica": 1},
         }
     }

oracle_ads-2.11.16/ads/aqua/config/deployment_config_defaults.json

@@ -0,0 +1,37 @@
+{
+  "configuration": {
+    "VM.Standard.A1.Flex": {
+      "parameters": {},
+      "shape_info": {
+        "configs": [
+          {
+            "memory_in_gbs": 128,
+            "ocpu": 20
+          },
+          {
+            "memory_in_gbs": 256,
+            "ocpu": 40
+          },
+          {
+            "memory_in_gbs": 384,
+            "ocpu": 60
+          },
+          {
+            "memory_in_gbs": 512,
+            "ocpu": 80
+          }
+        ],
+        "type": "CPU"
+      }
+    }
+  },
+  "shape": [
+    "VM.GPU.A10.1",
+    "VM.GPU.A10.2",
+    "BM.GPU.A10.4",
+    "BM.GPU4.8",
+    "BM.GPU.A100-v2.8",
+    "BM.GPU.H100.8",
+    "VM.Standard.A1.Flex"
+  ]
+}

{oracle_ads-2.11.14 → oracle_ads-2.11.16}/ads/aqua/config/resource_limit_names.json

@@ -1,6 +1,7 @@
 {
   "BM.GPU.A10.4": "ds-gpu-a10-count",
   "BM.GPU.A100-v2.8": "ds-gpu-a100-v2-count",
+  "BM.GPU.H100.8": "ds-gpu-h100-count",
   "BM.GPU4.8": "ds-gpu4-count",
   "VM.GPU.A10.1": "ds-gpu-a10-count",
   "VM.GPU.A10.2": "ds-gpu-a10-count"

{oracle_ads-2.11.14 → oracle_ads-2.11.16}/ads/aqua/constants.py

@@ -1,5 +1,4 @@
 #!/usr/bin/env python
-# -*- coding: utf-8 -*-
 # 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/
 """This module defines constants used in ads.aqua module."""
@@ -22,7 +21,6 @@ DEFAULT_FT_BLOCK_STORAGE_SIZE = 750
 DEFAULT_FT_REPLICA = 1
 DEFAULT_FT_BATCH_SIZE = 1
 DEFAULT_FT_VALIDATION_SET_SIZE = 0.1
-
 MAXIMUM_ALLOWED_DATASET_IN_BYTE = 52428800  # 1024 x 1024 x 50 = 50MB
 JOB_INFRASTRUCTURE_TYPE_DEFAULT_NETWORKING = "ME_STANDALONE"
 NB_SESSION_IDENTIFIER = "NB_SESSION_OCID"
@@ -35,6 +33,7 @@ AQUA_MODEL_TYPE_CUSTOM = "custom"
 AQUA_MODEL_ARTIFACT_CONFIG = "config.json"
 AQUA_MODEL_ARTIFACT_CONFIG_MODEL_NAME = "_name_or_path"
 AQUA_MODEL_ARTIFACT_CONFIG_MODEL_TYPE = "model_type"
+AQUA_MODEL_ARTIFACT_FILE = "model_file"
 
 TRAINING_METRICS_FINAL = "training_metrics_final"
 VALIDATION_METRICS_FINAL = "validation_metrics_final"
@@ -45,19 +44,37 @@ SERVICE_MANAGED_CONTAINER_URI_SCHEME = "dsmc://"
 SUPPORTED_FILE_FORMATS = ["jsonl"]
 MODEL_BY_REFERENCE_OSS_PATH_KEY = "artifact_location"
 
-CONSOLE_LINK_RESOURCE_TYPE_MAPPING = dict(
-    datasciencemodel="models",
-    datasciencemodeldeployment="model-deployments",
-    datasciencemodeldeploymentdev="model-deployments",
-    datasciencemodeldeploymentint="model-deployments",
-    datasciencemodeldeploymentpre="model-deployments",
-    datasciencejob="jobs",
-    datasciencejobrun="job-runs",
-    datasciencejobrundev="job-runs",
-    datasciencejobrunint="job-runs",
-    datasciencejobrunpre="job-runs",
-    datasciencemodelversionset="model-version-sets",
-    datasciencemodelversionsetpre="model-version-sets",
-    datasciencemodelversionsetint="model-version-sets",
-    datasciencemodelversionsetdev="model-version-sets",
-)
+CONSOLE_LINK_RESOURCE_TYPE_MAPPING = {
+    "datasciencemodel": "models",
+    "datasciencemodeldeployment": "model-deployments",
+    "datasciencemodeldeploymentdev": "model-deployments",
+    "datasciencemodeldeploymentint": "model-deployments",
+    "datasciencemodeldeploymentpre": "model-deployments",
+    "datasciencejob": "jobs",
+    "datasciencejobrun": "job-runs",
+    "datasciencejobrundev": "job-runs",
+    "datasciencejobrunint": "job-runs",
+    "datasciencejobrunpre": "job-runs",
+    "datasciencemodelversionset": "model-version-sets",
+    "datasciencemodelversionsetpre": "model-version-sets",
+    "datasciencemodelversionsetint": "model-version-sets",
+    "datasciencemodelversionsetdev": "model-version-sets",
+}
+
+VLLM_INFERENCE_RESTRICTED_PARAMS = {
+    "--port",
+    "--host",
+    "--served-model-name",
+    "--seed",
+}
+TGI_INFERENCE_RESTRICTED_PARAMS = {
+    "--port",
+    "--hostname",
+    "--num-shard",
+    "--sharded",
+    "--trust-remote-code",
+}
+LLAMA_CPP_INFERENCE_RESTRICTED_PARAMS = {
+    "--port",
+    "--host",
+}

{oracle_ads-2.11.14 → oracle_ads-2.11.16}/ads/aqua/evaluation/entities.py

@@ -1,5 +1,4 @@
 #!/usr/bin/env python
-# -*- coding: utf-8 -*-
 # 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/