aiauto-client 0.1.14__tar.gz → 0.1.16__tar.gz

{aiauto_client-0.1.14 → aiauto_client-0.1.16}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: aiauto-client
-Version: 0.1.14
+Version: 0.1.16
 Summary: AI Auto HPO (Hyperparameter Optimization) Client Library
 Author-email: AIAuto Team <ainode@zeroone.ai>
 Project-URL: Homepage, https://dashboard.common.aiauto.pangyo.ainode.ai
@@ -121,7 +121,7 @@ Jupyter Notebook이나 Python REPL에서 정의한 함수는 Serialize 할 수
 import aiauto
 import optuna
 
-def objective(trial: optuna.trial.Trial):
+def objective(trial):
     """
     이 함수는 외부 서버에서 실행됩니다.
     모든 import는 함수 내부에 작성하세요.
@@ -170,7 +170,7 @@ time.sleep(5)
 # `https://dashboard.common.aiauto.pangyo.ainode.ai/study` 에서 생성된 study 확인 가능
 
 # objective 함수 정의
-def objective(trial: optuna.trial.Trial):
+def objective(trial):
     """실제 실행은 사용자 로컬 컴퓨터가 아닌 서버에서 실행 될 함수"""
     x = trial.suggest_float('x', -10, 10)
     y = trial.suggest_float('y', -10, 10)
@@ -214,7 +214,7 @@ time.sleep(5)
 
 # objective 함수 정의
 # https://docs.pytorch.org/tutorials/beginner/blitz/cifar10_tutorial.html 참고
-def objective(trial: optuna.trial.Trial):
+def objective(trial):
     """
     실제 실행은 사용자 로컬 컴퓨터가 아닌 서버에서 실행 될 함수
     모든 import는 함수 내부에 존재해야 함
@@ -224,6 +224,8 @@ def objective(trial: optuna.trial.Trial):
     from torch.utils.data import DataLoader, random_split, Subset
     from torchvision import transforms, datasets
     import torch.nn.functional as F
+    
+    import optuna
 
     # 하이퍼파라미터 샘플링
     lr = trial.suggest_float('learning_rate', 1e-5, 1e-1, log=True)
@@ -315,7 +317,12 @@ study_wrapper.optimize(
     n_trials=100,
     parallelism=4,
     use_gpu=True,  # GPU 사용
-    requirements_list=['torch', 'torchvision']  # Pod에서 자동 설치
+    runtime_image='pytorch/pytorch:2.1.0-cuda11.8-cudnn8-runtime',  # default image for use_gpu True
+    # requirements_list=['torch', 'torchvision']  # Pod에서 자동 설치  # pip list 명시는 다운로드 받는데 느림, runtime_image 를 torch 로 명시하는게 나음
+    resources_requests={
+        "cpu": "2",
+        "memory": "4Gi",
+    },
 )
 time.sleep(5)
 ```
@@ -344,7 +351,7 @@ time.sleep(5)
 
 # objective 함수 정의
 # https://docs.pytorch.org/tutorials/beginner/blitz/cifar10_tutorial.html 참고
-def objective(trial: optuna.trial.Trial):
+def objective(trial):
     """
     실제 실행은 사용자 로컬 컴퓨터가 아닌 서버에서 실행 될 함수
     모든 import는 함수 내부에 존재해야 함
@@ -446,7 +453,16 @@ study_wrapper.optimize(
     n_trials=100,
     parallelism=4,
     use_gpu=True,  # GPU 사용
-    requirements_list=['torch', 'torchvision', 'fvcore']  # Pod에서 자동 설치
+    runtime_image='pytorch/pytorch:2.1.0-cuda11.8-cudnn8-runtime',  # default image for use_gpu True
+    requirements_list=[  # pip list 명시는 다운로드 받는데 느림, runtime_image 를 torch 로 명시하는게 나음
+        # 'torch', 
+        # 'torchvision', 
+        'fvcore',
+    ],  # Pod에서 자동 설치
+    resources_requests={
+        "cpu": "2",
+        "memory": "4Gi",
+    },
 )
 time.sleep(5)
 ```
@@ -545,15 +561,15 @@ study = controller.create_study('exp1', direction='minimize')
 time.sleep(5)
 
 def objective(trial):
-import numpy as np
-x = trial.suggest_float('x', -10, 10)
-return (x - 1.23) ** 2
+    import numpy as np
+    x = trial.suggest_float('x', -10, 10)
+    return (x - 1.23) ** 2
 
 study.optimize(
-objective,
-n_trials=64,
-parallelism=8,
-requirements_list=['numpy'],
+    objective,
+    n_trials=64,
+    parallelism=8,
+    requirements_list=['numpy'],
 )
 time.sleep(5)
 ```
@@ -600,17 +616,17 @@ import optuna, aiauto, time
 
 controller = aiauto.AIAutoController('aiauto_xxx')
 study = controller.create_study(
-study_name='cnn',
-direction='minimize',
-sampler=optuna.samplers.TPESampler(seed=42),
-pruner=optuna.pruners.MedianPruner(n_startup_trials=5),
+    study_name='cnn',
+    direction='minimize',
+    sampler=optuna.samplers.TPESampler(seed=42),
+    pruner=optuna.pruners.MedianPruner(n_startup_trials=5),
 )
 time.sleep(5)
 
 def objective(trial):
-import numpy as np
-lr = trial.suggest_float('lr', 1e-5, 1e-1, log=True)
-return (np.log10(lr) + 2) ** 2
+    import numpy as np
+    lr = trial.suggest_float('lr', 1e-5, 1e-1, log=True)
+    return (np.log10(lr) + 2) ** 2
 
 study.optimize(objective, n_trials=50, parallelism=4)
 time.sleep(5)

{aiauto_client-0.1.14 → aiauto_client-0.1.16}/README.md

@@ -94,7 +94,7 @@ Jupyter Notebook이나 Python REPL에서 정의한 함수는 Serialize 할 수
 import aiauto
 import optuna
 
-def objective(trial: optuna.trial.Trial):
+def objective(trial):
     """
     이 함수는 외부 서버에서 실행됩니다.
     모든 import는 함수 내부에 작성하세요.
@@ -143,7 +143,7 @@ time.sleep(5)
 # `https://dashboard.common.aiauto.pangyo.ainode.ai/study` 에서 생성된 study 확인 가능
 
 # objective 함수 정의
-def objective(trial: optuna.trial.Trial):
+def objective(trial):
     """실제 실행은 사용자 로컬 컴퓨터가 아닌 서버에서 실행 될 함수"""
     x = trial.suggest_float('x', -10, 10)
     y = trial.suggest_float('y', -10, 10)
@@ -187,7 +187,7 @@ time.sleep(5)
 
 # objective 함수 정의
 # https://docs.pytorch.org/tutorials/beginner/blitz/cifar10_tutorial.html 참고
-def objective(trial: optuna.trial.Trial):
+def objective(trial):
     """
     실제 실행은 사용자 로컬 컴퓨터가 아닌 서버에서 실행 될 함수
     모든 import는 함수 내부에 존재해야 함
@@ -197,6 +197,8 @@ def objective(trial: optuna.trial.Trial):
     from torch.utils.data import DataLoader, random_split, Subset
     from torchvision import transforms, datasets
     import torch.nn.functional as F
+    
+    import optuna
 
     # 하이퍼파라미터 샘플링
     lr = trial.suggest_float('learning_rate', 1e-5, 1e-1, log=True)
@@ -288,7 +290,12 @@ study_wrapper.optimize(
     n_trials=100,
     parallelism=4,
     use_gpu=True,  # GPU 사용
-    requirements_list=['torch', 'torchvision']  # Pod에서 자동 설치
+    runtime_image='pytorch/pytorch:2.1.0-cuda11.8-cudnn8-runtime',  # default image for use_gpu True
+    # requirements_list=['torch', 'torchvision']  # Pod에서 자동 설치  # pip list 명시는 다운로드 받는데 느림, runtime_image 를 torch 로 명시하는게 나음
+    resources_requests={
+        "cpu": "2",
+        "memory": "4Gi",
+    },
 )
 time.sleep(5)
 ```
@@ -317,7 +324,7 @@ time.sleep(5)
 
 # objective 함수 정의
 # https://docs.pytorch.org/tutorials/beginner/blitz/cifar10_tutorial.html 참고
-def objective(trial: optuna.trial.Trial):
+def objective(trial):
     """
     실제 실행은 사용자 로컬 컴퓨터가 아닌 서버에서 실행 될 함수
     모든 import는 함수 내부에 존재해야 함
@@ -419,7 +426,16 @@ study_wrapper.optimize(
     n_trials=100,
     parallelism=4,
     use_gpu=True,  # GPU 사용
-    requirements_list=['torch', 'torchvision', 'fvcore']  # Pod에서 자동 설치
+    runtime_image='pytorch/pytorch:2.1.0-cuda11.8-cudnn8-runtime',  # default image for use_gpu True
+    requirements_list=[  # pip list 명시는 다운로드 받는데 느림, runtime_image 를 torch 로 명시하는게 나음
+        # 'torch', 
+        # 'torchvision', 
+        'fvcore',
+    ],  # Pod에서 자동 설치
+    resources_requests={
+        "cpu": "2",
+        "memory": "4Gi",
+    },
 )
 time.sleep(5)
 ```
@@ -518,15 +534,15 @@ study = controller.create_study('exp1', direction='minimize')
 time.sleep(5)
 
 def objective(trial):
-import numpy as np
-x = trial.suggest_float('x', -10, 10)
-return (x - 1.23) ** 2
+    import numpy as np
+    x = trial.suggest_float('x', -10, 10)
+    return (x - 1.23) ** 2
 
 study.optimize(
-objective,
-n_trials=64,
-parallelism=8,
-requirements_list=['numpy'],
+    objective,
+    n_trials=64,
+    parallelism=8,
+    requirements_list=['numpy'],
 )
 time.sleep(5)
 ```
@@ -573,17 +589,17 @@ import optuna, aiauto, time
 
 controller = aiauto.AIAutoController('aiauto_xxx')
 study = controller.create_study(
-study_name='cnn',
-direction='minimize',
-sampler=optuna.samplers.TPESampler(seed=42),
-pruner=optuna.pruners.MedianPruner(n_startup_trials=5),
+    study_name='cnn',
+    direction='minimize',
+    sampler=optuna.samplers.TPESampler(seed=42),
+    pruner=optuna.pruners.MedianPruner(n_startup_trials=5),
 )
 time.sleep(5)
 
 def objective(trial):
-import numpy as np
-lr = trial.suggest_float('lr', 1e-5, 1e-1, log=True)
-return (np.log10(lr) + 2) ** 2
+    import numpy as np
+    lr = trial.suggest_float('lr', 1e-5, 1e-1, log=True)
+    return (np.log10(lr) + 2) ** 2
 
 study.optimize(objective, n_trials=50, parallelism=4)
 time.sleep(5)

{aiauto_client-0.1.14 → aiauto_client-0.1.16}/examples/example_torch_multiple_objective.py

@@ -49,6 +49,7 @@ def objective_multi(trial):
     from fvcore.nn import FlopCountAnalysis
 
     from optuna.artifacts import upload_artifact
+    import aiauto
 
 
     # objective 함수의 매개변수로 받아온 optuna 자체의 trial 을 aiauto 에서 사용하는 TrialController 로 Warpping Log 찍는 용도
@@ -259,6 +260,7 @@ def objective_detailed(trial):
     from fvcore.nn import FlopCountAnalysis
 
     from optuna.artifacts import upload_artifact
+    import aiauto
 
 
     # objective 함수의 매개변수로 받아온 optuna 자체의 trial 을 aiauto 에서 사용하는 TrialController 로 Warpping Log 찍는 용도
@@ -440,14 +442,18 @@ if __name__ == '__main__':
         n_trials=100,
         parallelism=4,  # n_jobs 대신 parallelism 사용
         use_gpu=True,  # GPU 사용
-        # runtime_image = "pytorch/pytorch:2.1.0-cuda12.1-cudnn8-runtime",  # default image for use_gpu True
+        runtime_image='pytorch/pytorch:2.1.0-cuda11.8-cudnn8-runtime',  # default image for use_gpu True
         # requirements.txt 대신 리스트로 전달 (외부 서버에서 설치)
-        requirements_list=[
-            'torch',
-            'torchvision',
+        requirements_list=[  # pip list 명시는 다운로드 받는데 느림, runtime_image 를 torch 로 명시하는게 나음
+            # 'torch',
+            # 'torchvision',
             'fvcore',  # FLOPS 계산용 (multi-objective 사용 시)
         ],
         # CallbackTopNArtifact은 클라이언트 측에서는 사용 불가 (runner 에서 자동으로 지정 됨)
+        resources_requests={
+            "cpu": "2",
+            "memory": "4Gi",
+        },
     )
     time.sleep(5)
 
@@ -470,4 +476,8 @@ if __name__ == '__main__':
         use_gpu=True,
         # runtime_image = "pytorch/pytorch:2.1.0-cuda12.1-cudnn8-runtime",  # default image for use_gpu True
         requirements_list=['torch', 'torchvision', 'fvcore'],
+        resources_requests={
+            "cpu": "2",
+            "memory": "4Gi",
+        },
     )

{aiauto_client-0.1.14 → aiauto_client-0.1.16}/examples/example_torch_single_objective.py

@@ -47,7 +47,9 @@ def objective_single(trial):
     from torchvision import transforms, datasets
     import torch.nn.functional as F
 
+    import optuna
     from optuna.artifacts import upload_artifact
+    import aiauto
 
 
     # objective 함수의 매개변수로 받아온 optuna 자체의 trial 을 aiauto 에서 사용하는 TrialController 로 Warpping Log 찍는 용도
@@ -261,6 +263,7 @@ def objective_detailed(trial):
     import torch.nn.functional as F
 
     from optuna.artifacts import upload_artifact
+    import aiauto
 
 
     # objective 함수의 매개변수로 받아온 optuna 자체의 trial 을 aiauto 에서 사용하는 TrialController 로 Warpping Log 찍는 용도
@@ -447,10 +450,14 @@ if __name__ == '__main__':
         n_trials=100,
         parallelism=4,  # n_jobs 대신 parallelism 사용
         use_gpu=True,  # GPU 사용
-        # runtime_image = "pytorch/pytorch:2.1.0-cuda12.1-cudnn8-runtime",  # default image for use_gpu True
+        runtime_image='pytorch/pytorch:2.1.0-cuda11.8-cudnn8-runtime',  # default image for use_gpu True
         # objective 함수 안에서 import 하는거 requirements.txt 대신 리스트로 전달 (pod 안에서 설치)
-        requirements_list=['torch', 'torchvision'],
+        # requirements_list=['torch', 'torchvision'],  # pip list 명시는 다운로드 받는데 느림, runtime_image 를 torch 로 명시하는게 나음
         # CallbackTopNArtifact은 클라이언트 측에서는 사용 불가 (runner 에서 자동으로 지정 됨)
+        resources_requests={
+            "cpu": "2",
+            "memory": "4Gi",
+        },
     )
     time.sleep(5)
 
@@ -473,4 +480,8 @@ if __name__ == '__main__':
         use_gpu=True,
         # runtime_image = "pytorch/pytorch:2.1.0-cuda12.1-cudnn8-runtime",  # default image for use_gpu True
         requirements_list=['torch', 'torchvision'],
+        resources_requests={
+            "cpu": "2",
+            "memory": "4Gi",
+        },
     )

{aiauto_client-0.1.14 → aiauto_client-0.1.16}/examples/simple_example.py

@@ -26,7 +26,7 @@ ac = aiauto.AIAutoController('<token>')
 
 
 # single objective accuracy
-def simple_objective(trial):
+def objective_simple(trial):
     import optuna
     import aiauto
     """
@@ -74,7 +74,7 @@ if __name__ == '__main__':
         # runtime_image = "ghcr.io/astral-sh/uv:python3.8-bookworm-slim",  # default image for use_gpu False
     )
     time.sleep(5)
-    
+
     study = study_wrapper.get_study()
 
     print('\nBest trials:')

{aiauto_client-0.1.14 → aiauto_client-0.1.16}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "aiauto-client"
-version = "0.1.14"
+version = "0.1.16"
 description = "AI Auto HPO (Hyperparameter Optimization) Client Library"
 readme = "README.md"
 requires-python = ">=3.8"

{aiauto_client-0.1.14 → aiauto_client-0.1.16}/src/aiauto/core.py

@@ -188,11 +188,24 @@ class StudyWrapper:
         parallelism: int = 2,
         requirements_file: Optional[str] = None,
         requirements_list: Optional[List[str]] = None,
-        resources_requests: Optional[Dict[str, str]] = {"cpu": "256m", "memory": "256Mi"},
-        resources_limits: Optional[Dict[str, str]] = {"cpu": "256m", "memory": "256Mi"},
+        resources_requests: Optional[Dict[str, str]] = None,
+        resources_limits: Optional[Dict[str, str]] = None,
         runtime_image: Optional[str] = 'ghcr.io/astral-sh/uv:python3.8-bookworm-slim',
         use_gpu: bool = False
     ) -> None:
+        # 리소스 기본값 설정
+        if resources_requests is None:
+            if use_gpu:
+                resources_requests = {"cpu": "2", "memory": "4Gi"}
+            else:
+                resources_requests = {"cpu": "1", "memory": "1Gi"}
+        
+        if resources_limits is None:
+            if use_gpu:
+                resources_limits = {"cpu": "2", "memory": "4Gi"}
+            else:
+                resources_limits = {"cpu": "1", "memory": "1Gi"}
+        
         if runtime_image is None or runtime_image == "":
             if use_gpu:
                 runtime_image = "pytorch/pytorch:2.1.0-cuda12.1-cudnn8-runtime"

{aiauto_client-0.1.14 → aiauto_client-0.1.16}/src/aiauto/serializer.py

@@ -59,9 +59,12 @@ def object_to_json(obj: Union[object, dict, None]) -> str:
             
             # __init__의 실제 파라미터인지 확인
             if param_name in valid_params:
-                # PatientPruner의 wrapped_pruner만 특별 처리
+                # PatientPruner의 wrapped_pruner 특별 처리
                 if class_name == "PatientPruner" and param_name == "wrapped_pruner" and value is not None:
                     kwargs[param_name] = json.loads(object_to_json(value))
+                # CmaEsSampler와 QMCSampler의 independent_sampler 특별 처리
+                elif param_name == "independent_sampler" and value is not None and class_name in ["CmaEsSampler", "QMCSampler"]:
+                    kwargs[param_name] = json.loads(object_to_json(value))
                 # Callable 타입은 제외 (gamma, weights 등)
                 elif not callable(value):
                     kwargs[param_name] = value

{aiauto_client-0.1.14 → aiauto_client-0.1.16}/src/aiauto_client.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: aiauto-client
-Version: 0.1.14
+Version: 0.1.16
 Summary: AI Auto HPO (Hyperparameter Optimization) Client Library
 Author-email: AIAuto Team <ainode@zeroone.ai>
 Project-URL: Homepage, https://dashboard.common.aiauto.pangyo.ainode.ai
@@ -121,7 +121,7 @@ Jupyter Notebook이나 Python REPL에서 정의한 함수는 Serialize 할 수
 import aiauto
 import optuna
 
-def objective(trial: optuna.trial.Trial):
+def objective(trial):
     """
     이 함수는 외부 서버에서 실행됩니다.
     모든 import는 함수 내부에 작성하세요.
@@ -170,7 +170,7 @@ time.sleep(5)
 # `https://dashboard.common.aiauto.pangyo.ainode.ai/study` 에서 생성된 study 확인 가능
 
 # objective 함수 정의
-def objective(trial: optuna.trial.Trial):
+def objective(trial):
     """실제 실행은 사용자 로컬 컴퓨터가 아닌 서버에서 실행 될 함수"""
     x = trial.suggest_float('x', -10, 10)
     y = trial.suggest_float('y', -10, 10)
@@ -214,7 +214,7 @@ time.sleep(5)
 
 # objective 함수 정의
 # https://docs.pytorch.org/tutorials/beginner/blitz/cifar10_tutorial.html 참고
-def objective(trial: optuna.trial.Trial):
+def objective(trial):
     """
     실제 실행은 사용자 로컬 컴퓨터가 아닌 서버에서 실행 될 함수
     모든 import는 함수 내부에 존재해야 함
@@ -224,6 +224,8 @@ def objective(trial: optuna.trial.Trial):
     from torch.utils.data import DataLoader, random_split, Subset
     from torchvision import transforms, datasets
     import torch.nn.functional as F
+    
+    import optuna
 
     # 하이퍼파라미터 샘플링
     lr = trial.suggest_float('learning_rate', 1e-5, 1e-1, log=True)
@@ -315,7 +317,12 @@ study_wrapper.optimize(
     n_trials=100,
     parallelism=4,
     use_gpu=True,  # GPU 사용
-    requirements_list=['torch', 'torchvision']  # Pod에서 자동 설치
+    runtime_image='pytorch/pytorch:2.1.0-cuda11.8-cudnn8-runtime',  # default image for use_gpu True
+    # requirements_list=['torch', 'torchvision']  # Pod에서 자동 설치  # pip list 명시는 다운로드 받는데 느림, runtime_image 를 torch 로 명시하는게 나음
+    resources_requests={
+        "cpu": "2",
+        "memory": "4Gi",
+    },
 )
 time.sleep(5)
 ```
@@ -344,7 +351,7 @@ time.sleep(5)
 
 # objective 함수 정의
 # https://docs.pytorch.org/tutorials/beginner/blitz/cifar10_tutorial.html 참고
-def objective(trial: optuna.trial.Trial):
+def objective(trial):
     """
     실제 실행은 사용자 로컬 컴퓨터가 아닌 서버에서 실행 될 함수
     모든 import는 함수 내부에 존재해야 함
@@ -446,7 +453,16 @@ study_wrapper.optimize(
     n_trials=100,
     parallelism=4,
     use_gpu=True,  # GPU 사용
-    requirements_list=['torch', 'torchvision', 'fvcore']  # Pod에서 자동 설치
+    runtime_image='pytorch/pytorch:2.1.0-cuda11.8-cudnn8-runtime',  # default image for use_gpu True
+    requirements_list=[  # pip list 명시는 다운로드 받는데 느림, runtime_image 를 torch 로 명시하는게 나음
+        # 'torch', 
+        # 'torchvision', 
+        'fvcore',
+    ],  # Pod에서 자동 설치
+    resources_requests={
+        "cpu": "2",
+        "memory": "4Gi",
+    },
 )
 time.sleep(5)
 ```
@@ -545,15 +561,15 @@ study = controller.create_study('exp1', direction='minimize')
 time.sleep(5)
 
 def objective(trial):
-import numpy as np
-x = trial.suggest_float('x', -10, 10)
-return (x - 1.23) ** 2
+    import numpy as np
+    x = trial.suggest_float('x', -10, 10)
+    return (x - 1.23) ** 2
 
 study.optimize(
-objective,
-n_trials=64,
-parallelism=8,
-requirements_list=['numpy'],
+    objective,
+    n_trials=64,
+    parallelism=8,
+    requirements_list=['numpy'],
 )
 time.sleep(5)
 ```
@@ -600,17 +616,17 @@ import optuna, aiauto, time
 
 controller = aiauto.AIAutoController('aiauto_xxx')
 study = controller.create_study(
-study_name='cnn',
-direction='minimize',
-sampler=optuna.samplers.TPESampler(seed=42),
-pruner=optuna.pruners.MedianPruner(n_startup_trials=5),
+    study_name='cnn',
+    direction='minimize',
+    sampler=optuna.samplers.TPESampler(seed=42),
+    pruner=optuna.pruners.MedianPruner(n_startup_trials=5),
 )
 time.sleep(5)
 
 def objective(trial):
-import numpy as np
-lr = trial.suggest_float('lr', 1e-5, 1e-1, log=True)
-return (np.log10(lr) + 2) ** 2
+    import numpy as np
+    lr = trial.suggest_float('lr', 1e-5, 1e-1, log=True)
+    return (np.log10(lr) + 2) ** 2
 
 study.optimize(objective, n_trials=50, parallelism=4)
 time.sleep(5)

{aiauto_client-0.1.14 → aiauto_client-0.1.16}/src/aiauto_client.egg-info/SOURCES.txt

@@ -15,4 +15,5 @@ src/aiauto_client.egg-info/SOURCES.txt
 src/aiauto_client.egg-info/dependency_links.txt
 src/aiauto_client.egg-info/requires.txt
 src/aiauto_client.egg-info/top_level.txt
-tests/test_local_storage.py
+tests/test_pruners.py
+tests/test_samplers.py

aiauto_client-0.1.16/tests/test_pruners.py

@@ -0,0 +1,205 @@
+#!/usr/bin/env python
+"""
+Optuna Pruner 직렬화/역직렬화 테스트
+
+모든 Optuna Pruner가 올바르게 직렬화되고 복원되는지 검증합니다.
+특히 PatientPruner의 wrapped_pruner 재귀 처리를 중점적으로 테스트합니다.
+"""
+
+import sys
+import os
+import json
+import optuna
+import pytest
+
+# Add parent directory to path for imports
+sys.path.insert(0, os.path.join(os.path.dirname(__file__), '..', 'src'))
+sys.path.insert(0, os.path.join(os.path.dirname(__file__), '../../runners'))
+
+from aiauto.serializer import object_to_json
+from runner_create_study import from_json, PRUNER_WHITELIST
+
+
+def test_nop_pruner():
+    """NopPruner 테스트 - 파라미터 없음"""
+    pruner = optuna.pruners.NopPruner()
+    json_str = object_to_json(pruner)
+    
+    # JSON 파싱 가능 여부 확인
+    parsed = json.loads(json_str)
+    assert parsed["cls"] == "NopPruner"
+    # NopPruner는 파라미터가 없으므로 args가 비어있거나 없을 수 있음
+    
+    # 역직렬화
+    restored = from_json(json_str, PRUNER_WHITELIST)
+    assert type(restored).__name__ == "NopPruner"
+
+
+def test_median_pruner():
+    """MedianPruner 테스트 - 기본 파라미터"""
+    pruner = optuna.pruners.MedianPruner(
+        n_startup_trials=5,
+        n_warmup_steps=10,
+        interval_steps=2
+    )
+    json_str = object_to_json(pruner)
+    
+    parsed = json.loads(json_str)
+    assert parsed["cls"] == "MedianPruner"
+    assert parsed["kwargs"]["n_startup_trials"] == 5
+    assert parsed["kwargs"]["n_warmup_steps"] == 10
+    assert parsed["kwargs"]["interval_steps"] == 2
+    
+    restored = from_json(json_str, PRUNER_WHITELIST)
+    assert type(restored).__name__ == "MedianPruner"
+    assert restored._n_startup_trials == 5
+    assert restored._n_warmup_steps == 10
+    assert restored._interval_steps == 2
+
+
+def test_percentile_pruner():
+    """PercentilePruner 테스트 - percentile 필수 파라미터"""
+    pruner = optuna.pruners.PercentilePruner(
+        percentile=25.0,
+        n_startup_trials=3
+    )
+    json_str = object_to_json(pruner)
+    
+    parsed = json.loads(json_str)
+    assert parsed["cls"] == "PercentilePruner"
+    assert parsed["kwargs"]["percentile"] == 25.0
+    assert parsed["kwargs"]["n_startup_trials"] == 3
+    
+    restored = from_json(json_str, PRUNER_WHITELIST)
+    assert type(restored).__name__ == "PercentilePruner"
+    assert restored._percentile == 25.0
+
+
+def test_patient_pruner_with_wrapped():
+    """PatientPruner 테스트 - wrapped_pruner 재귀 처리"""
+    wrapped = optuna.pruners.MedianPruner(n_startup_trials=10)
+    pruner = optuna.pruners.PatientPruner(
+        wrapped_pruner=wrapped,
+        patience=3,
+        min_delta=0.1
+    )
+    json_str = object_to_json(pruner)
+    
+    parsed = json.loads(json_str)
+    assert parsed["cls"] == "PatientPruner"
+    assert parsed["kwargs"]["patience"] == 3
+    assert parsed["kwargs"]["min_delta"] == 0.1
+    
+    # wrapped_pruner가 재귀적으로 직렬화되었는지 확인
+    assert "wrapped_pruner" in parsed["kwargs"]
+    assert parsed["kwargs"]["wrapped_pruner"]["cls"] == "MedianPruner"
+    assert parsed["kwargs"]["wrapped_pruner"]["kwargs"]["n_startup_trials"] == 10
+    
+    restored = from_json(json_str, PRUNER_WHITELIST)
+    assert type(restored).__name__ == "PatientPruner"
+    assert restored._patience == 3
+    assert restored._min_delta == 0.1
+    assert type(restored._wrapped_pruner).__name__ == "MedianPruner"
+    assert restored._wrapped_pruner._n_startup_trials == 10
+
+
+def test_patient_pruner_without_wrapped():
+    """PatientPruner 테스트 - wrapped_pruner None으로"""
+    pruner = optuna.pruners.PatientPruner(wrapped_pruner=None, patience=5)
+    json_str = object_to_json(pruner)
+    
+    parsed = json.loads(json_str)
+    assert parsed["cls"] == "PatientPruner"
+    assert parsed["kwargs"]["patience"] == 5
+    # wrapped_pruner가 None이면 직렬화에서 제외될 수 있음
+    if "wrapped_pruner" in parsed["kwargs"]:
+        assert parsed["kwargs"]["wrapped_pruner"] is None
+    
+    restored = from_json(json_str, PRUNER_WHITELIST)
+    assert type(restored).__name__ == "PatientPruner"
+    assert restored._patience == 5
+    assert restored._wrapped_pruner is None
+
+
+def test_threshold_pruner():
+    """ThresholdPruner 테스트"""
+    pruner = optuna.pruners.ThresholdPruner(
+        lower=0.1,
+        upper=0.9,
+        n_warmup_steps=5
+    )
+    json_str = object_to_json(pruner)
+    
+    parsed = json.loads(json_str)
+    assert parsed["cls"] == "ThresholdPruner"
+    assert parsed["kwargs"]["lower"] == 0.1
+    assert parsed["kwargs"]["upper"] == 0.9
+    
+    restored = from_json(json_str, PRUNER_WHITELIST)
+    assert type(restored).__name__ == "ThresholdPruner"
+    assert restored._lower == 0.1
+    assert restored._upper == 0.9
+
+
+def test_successive_halving_pruner():
+    """SuccessiveHalvingPruner 테스트"""
+    pruner = optuna.pruners.SuccessiveHalvingPruner(
+        min_resource=1,
+        reduction_factor=4,
+        min_early_stopping_rate=0
+    )
+    json_str = object_to_json(pruner)
+    
+    parsed = json.loads(json_str)
+    assert parsed["cls"] == "SuccessiveHalvingPruner"
+    
+    restored = from_json(json_str, PRUNER_WHITELIST)
+    assert type(restored).__name__ == "SuccessiveHalvingPruner"
+
+
+def test_hyperband_pruner():
+    """HyperbandPruner 테스트"""
+    pruner = optuna.pruners.HyperbandPruner(
+        min_resource=1,
+        max_resource=100,
+        reduction_factor=3
+    )
+    json_str = object_to_json(pruner)
+    
+    parsed = json.loads(json_str)
+    assert parsed["cls"] == "HyperbandPruner"
+    assert parsed["kwargs"]["min_resource"] == 1
+    assert parsed["kwargs"]["max_resource"] == 100
+    assert parsed["kwargs"]["reduction_factor"] == 3
+    
+    restored = from_json(json_str, PRUNER_WHITELIST)
+    assert type(restored).__name__ == "HyperbandPruner"
+
+
+if __name__ == "__main__":
+    # 각 테스트 실행
+    test_nop_pruner()
+    print("✅ NopPruner test passed")
+    
+    test_median_pruner()
+    print("✅ MedianPruner test passed")
+    
+    test_percentile_pruner()
+    print("✅ PercentilePruner test passed")
+    
+    test_patient_pruner_with_wrapped()
+    print("✅ PatientPruner with wrapped test passed")
+    
+    test_patient_pruner_without_wrapped()
+    print("✅ PatientPruner without wrapped test passed")
+    
+    test_threshold_pruner()
+    print("✅ ThresholdPruner test passed")
+    
+    test_successive_halving_pruner()
+    print("✅ SuccessiveHalvingPruner test passed")
+    
+    test_hyperband_pruner()
+    print("✅ HyperbandPruner test passed")
+    
+    print("\n✅ All Pruner tests passed!")

aiauto_client-0.1.16/tests/test_samplers.py

@@ -0,0 +1,226 @@
+#!/usr/bin/env python
+"""
+Optuna Sampler 직렬화/역직렬화 테스트
+
+주요 Optuna Sampler가 올바르게 직렬화되고 복원되는지 검증합니다.
+Callable 타입 파라미터(gamma, weights 등)는 직렬화에서 제외됨을 확인합니다.
+"""
+
+import sys
+import os
+import json
+import optuna
+import pytest
+
+# Add parent directory to path for imports
+sys.path.insert(0, os.path.join(os.path.dirname(__file__), '..', 'src'))
+sys.path.insert(0, os.path.join(os.path.dirname(__file__), '../../runners'))
+
+from aiauto.serializer import object_to_json
+from runner_create_study import from_json, SAMPLER_WHITELIST
+
+
+def test_random_sampler():
+    """RandomSampler 테스트 - seed 파라미터"""
+    sampler = optuna.samplers.RandomSampler(seed=42)
+    json_str = object_to_json(sampler)
+    
+    parsed = json.loads(json_str)
+    assert parsed["cls"] == "RandomSampler"
+    # seed는 저장되지만 다른 형태로 저장될 수 있음
+    if "seed" in parsed["kwargs"]:
+        assert parsed["kwargs"]["seed"] == 42
+    
+    restored = from_json(json_str, SAMPLER_WHITELIST)
+    assert type(restored).__name__ == "RandomSampler"
+
+
+def test_tpe_sampler():
+    """TPESampler 테스트 - 주요 파라미터"""
+    sampler = optuna.samplers.TPESampler(
+        n_startup_trials=5,
+        n_ei_candidates=10,
+        seed=42,
+        multivariate=True,
+        warn_independent_sampling=False
+    )
+    json_str = object_to_json(sampler)
+    
+    parsed = json.loads(json_str)
+    assert parsed["cls"] == "TPESampler"
+    assert parsed["kwargs"]["n_startup_trials"] == 5
+    assert parsed["kwargs"]["n_ei_candidates"] == 10
+    assert parsed["kwargs"]["multivariate"] == True
+    assert parsed["kwargs"]["warn_independent_sampling"] == False
+    if "seed" in parsed["kwargs"]:
+        assert parsed["kwargs"]["seed"] == 42
+    
+    # Callable 파라미터들(gamma, weights)은 제외되어야 함
+    assert "gamma" not in parsed["kwargs"]
+    assert "weights" not in parsed["kwargs"]
+    
+    restored = from_json(json_str, SAMPLER_WHITELIST)
+    assert type(restored).__name__ == "TPESampler"
+    assert restored._n_startup_trials == 5
+    assert restored._n_ei_candidates == 10
+    if hasattr(restored, '_seed'):
+        assert restored._seed == 42
+
+
+def test_tpe_sampler_default():
+    """TPESampler 테스트 - 기본값"""
+    sampler = optuna.samplers.TPESampler()
+    json_str = object_to_json(sampler)
+    
+    parsed = json.loads(json_str)
+    assert parsed["cls"] == "TPESampler"
+    
+    # 기본값들이 올바르게 직렬화되는지 확인
+    assert "n_startup_trials" in parsed["kwargs"]
+    assert "n_ei_candidates" in parsed["kwargs"]
+    
+    restored = from_json(json_str, SAMPLER_WHITELIST)
+    assert type(restored).__name__ == "TPESampler"
+
+
+def test_nsgaii_sampler():
+    """NSGAIISampler 테스트 - 다목적 최적화용"""
+    sampler = optuna.samplers.NSGAIISampler(
+        population_size=10,
+        mutation_prob=0.1,
+        crossover_prob=0.8,
+        swapping_prob=0.4,
+        seed=42
+    )
+    json_str = object_to_json(sampler)
+    
+    parsed = json.loads(json_str)
+    assert parsed["cls"] == "NSGAIISampler"
+    assert parsed["kwargs"]["population_size"] == 10
+    if "mutation_prob" in parsed["kwargs"]:
+        assert parsed["kwargs"]["mutation_prob"] == 0.1
+    if "crossover_prob" in parsed["kwargs"]:
+        assert parsed["kwargs"]["crossover_prob"] == 0.8
+    if "swapping_prob" in parsed["kwargs"]:
+        assert parsed["kwargs"]["swapping_prob"] == 0.4
+    if "seed" in parsed["kwargs"]:
+        assert parsed["kwargs"]["seed"] == 42
+    
+    # Callable 파라미터들은 제외되거나 null이어야 함
+    assert parsed["kwargs"].get("constraints_func") is None
+    assert "elite_population_selection_strategy" not in parsed["kwargs"]
+    
+    restored = from_json(json_str, SAMPLER_WHITELIST)
+    assert type(restored).__name__ == "NSGAIISampler"
+    assert restored._population_size == 10
+    if hasattr(restored, '_seed'):
+        assert restored._seed == 42
+
+
+def test_grid_sampler():
+    """GridSampler 테스트 - search_space 파라미터"""
+    search_space = {
+        "x": [-10, -5, 0, 5, 10],
+        "y": [1, 2, 3]
+    }
+    sampler = optuna.samplers.GridSampler(search_space=search_space)
+    json_str = object_to_json(sampler)
+    
+    parsed = json.loads(json_str)
+    assert parsed["cls"] == "GridSampler"
+    assert parsed["kwargs"]["search_space"] == search_space
+    
+    restored = from_json(json_str, SAMPLER_WHITELIST)
+    assert type(restored).__name__ == "GridSampler"
+    assert restored._search_space == search_space
+
+
+def test_bruteforce_sampler():
+    """BruteForceSampler 테스트"""
+    sampler = optuna.samplers.BruteForceSampler(seed=42)
+    json_str = object_to_json(sampler)
+    
+    parsed = json.loads(json_str)
+    assert parsed["cls"] == "BruteForceSampler"
+    if "seed" in parsed["kwargs"]:
+        assert parsed["kwargs"]["seed"] == 42
+    
+    restored = from_json(json_str, SAMPLER_WHITELIST)
+    assert type(restored).__name__ == "BruteForceSampler"
+    if hasattr(restored, '_seed'):
+        assert restored._seed == 42
+
+
+def test_cmaes_sampler():
+    """CmaEsSampler 테스트"""
+    sampler = optuna.samplers.CmaEsSampler(
+        n_startup_trials=5,
+        seed=42
+    )
+    json_str = object_to_json(sampler)
+    
+    parsed = json.loads(json_str)
+    assert parsed["cls"] == "CmaEsSampler"
+    assert parsed["kwargs"]["n_startup_trials"] == 5
+    if "seed" in parsed["kwargs"]:
+        assert parsed["kwargs"]["seed"] == 42
+    # restart_strategy는 deprecated되어 저장되지 않음
+    
+    restored = from_json(json_str, SAMPLER_WHITELIST)
+    assert type(restored).__name__ == "CmaEsSampler"
+    assert restored._n_startup_trials == 5
+    if hasattr(restored, '_seed'):
+        assert restored._seed == 42
+
+
+def test_qmc_sampler():
+    """QMCSampler 테스트"""
+    sampler = optuna.samplers.QMCSampler(
+        qmc_type="sobol",
+        scramble=True,
+        seed=42
+    )
+    json_str = object_to_json(sampler)
+    
+    parsed = json.loads(json_str)
+    assert parsed["cls"] == "QMCSampler"
+    assert parsed["kwargs"]["qmc_type"] == "sobol"
+    assert parsed["kwargs"]["scramble"] == True
+    if "seed" in parsed["kwargs"]:
+        assert parsed["kwargs"]["seed"] == 42
+    
+    restored = from_json(json_str, SAMPLER_WHITELIST)
+    assert type(restored).__name__ == "QMCSampler"
+    assert restored._qmc_type == "sobol"
+    assert restored._scramble == True
+    if hasattr(restored, '_seed'):
+        assert restored._seed == 42
+
+
+if __name__ == "__main__":
+    # 각 테스트 실행
+    test_random_sampler()
+    print("✅ RandomSampler test passed")
+    
+    test_tpe_sampler()
+    print("✅ TPESampler test passed")
+    
+    test_tpe_sampler_default()
+    print("✅ TPESampler (default) test passed")
+    
+    test_nsgaii_sampler()
+    print("✅ NSGAIISampler test passed")
+    
+    test_grid_sampler()
+    print("✅ GridSampler test passed")
+    
+    test_bruteforce_sampler()
+    print("✅ BruteForceSampler test passed")
+    
+    test_cmaes_sampler()
+    print("✅ CmaEsSampler test passed")
+    
+    test_qmc_sampler()
+    print("✅ QMCSampler test passed")
+    
+    print("\n✅ All Sampler tests passed!")

aiauto_client-0.1.14/tests/test_local_storage.py

@@ -1,59 +0,0 @@
-import aiauto
-import optuna
-from unittest.mock import patch
-
-
-def objective(trial):
-    """간단한 이차함수 최적화 예제"""
-    # TrialController를 사용한 로깅
-    tc = aiauto.TrialController(trial)
-    tc.log("Starting simple optimization example")
-
-    # 하이퍼파라미터 샘플링
-    x = trial.suggest_float('x', -10, 10)
-    y = trial.suggest_float('y', -10, 10)
-
-    # 목적함수: (x-2)² + (y-5)² 최소화
-    result = (x - 2) ** 2 + (y - 5) ** 2
-
-    tc.log(f"x={x:.3f}, y={y:.3f}, result={result:.3f}")
-
-    return result
-
-
-def main():
-    print("🚀 AIAuto 소스코드 직렬화 로컬 테스트")
-
-    # AIAutoController의 storage를 InMemoryStorage로 패치
-    with patch.object(aiauto.AIAutoController, '__init__', lambda self: None):
-        controller = aiauto.AIAutoController()
-        # 로컬 테스트용 InMemoryStorage 설정
-        controller.storage = optuna.storages.InMemoryStorage()
-        controller.artifact_store = optuna.artifacts.FileSystemArtifactStore('./artifacts')
-
-        # 소스코드 직렬화 테스트
-        print("\n=== 소스코드 직렬화 테스트 ===")
-        study_wrapper = controller.create_study(
-            objective=objective,
-            study_name='local_test',
-            direction='minimize'
-        )
-
-        print("✅ StudyWrapper 생성 성공!")
-
-        # 최적화 실행
-        print("\n=== 최적화 실행 ===")
-        study_wrapper.optimize(n_trials=10)
-
-        # 결과 출력
-        print(f"\n🎉 최적화 완료!")
-        print(f"📊 Best value: {study_wrapper.best_value:.3f}")
-        print(f"🔧 Best params: {study_wrapper.best_params}")
-
-        # 이론적 최적해: x=2, y=5, result=0
-        print(f"💡 이론적 최적해: x=2, y=5, result=0")
-        print(f"📈 오차: {study_wrapper.best_value:.3f}")
-
-
-if __name__ == "__main__":
-    main()