feat: refactor CoSTEER classes to use DSCoSTEER and update max seconds handling (#1156)

* feat: refactor CoSTEER classes to use DSCoSTEER and update max seconds handling

* remove useless line

* enable time_ratio_limit_to_enable_hyperparameter_tuning

rdagent/components/coder/data_science/ensemble/__init__.py

@@ -16,7 +16,6 @@ from pathlib import Path
 from jinja2 import Environment, StrictUndefined
 
 from rdagent.app.data_science.conf import DS_RD_SETTING
-from rdagent.components.coder.CoSTEER import CoSTEER
 from rdagent.components.coder.CoSTEER.evaluators import (
     CoSTEERMultiEvaluator,
     CoSTEERSingleFeedback,
@@ -30,6 +29,7 @@ from rdagent.components.coder.CoSTEER.knowledge_management import (
 from rdagent.components.coder.data_science.conf import DSCoderCoSTEERSettings
 from rdagent.components.coder.data_science.ensemble.eval import EnsembleCoSTEEREvaluator
 from rdagent.components.coder.data_science.ensemble.exp import EnsembleTask
+from rdagent.components.coder.data_science.share.ds_costeer import DSCoSTEER
 from rdagent.core.exception import CoderError
 from rdagent.core.experiment import FBWorkspace
 from rdagent.core.scenario import Scenario
@@ -141,7 +141,7 @@ class EnsembleMultiProcessEvolvingStrategy(MultiProcessEvolvingStrategy):
         return evo
 
 
-class EnsembleCoSTEER(CoSTEER):
+class EnsembleCoSTEER(DSCoSTEER):
     def __init__(
         self,
         scen: Scenario,
@@ -160,6 +160,5 @@ class EnsembleCoSTEER(CoSTEER):
             evolving_version=2,
             scen=scen,
             max_loop=DS_RD_SETTING.coder_max_loop,
-            max_seconds=scen.real_debug_timeout() * settings.max_seconds_multiplier,
             **kwargs,
         )

rdagent/components/coder/data_science/feature/__init__.py

@@ -1,9 +1,6 @@
-import json
 from pathlib import Path
-from typing import Dict
 
 from rdagent.app.data_science.conf import DS_RD_SETTING
-from rdagent.components.coder.CoSTEER import CoSTEER
 from rdagent.components.coder.CoSTEER.evaluators import (
     CoSTEERMultiEvaluator,
     CoSTEERSingleFeedback,
@@ -17,6 +14,7 @@ from rdagent.components.coder.CoSTEER.knowledge_management import (
 from rdagent.components.coder.data_science.conf import DSCoderCoSTEERSettings
 from rdagent.components.coder.data_science.feature.eval import FeatureCoSTEEREvaluator
 from rdagent.components.coder.data_science.feature.exp import FeatureTask
+from rdagent.components.coder.data_science.share.ds_costeer import DSCoSTEER
 from rdagent.core.exception import CoderError
 from rdagent.core.experiment import FBWorkspace
 from rdagent.core.scenario import Scenario
@@ -117,7 +115,7 @@ class FeatureMultiProcessEvolvingStrategy(MultiProcessEvolvingStrategy):
         return evo
 
 
-class FeatureCoSTEER(CoSTEER):
+class FeatureCoSTEER(DSCoSTEER):
     def __init__(
         self,
         scen: Scenario,
@@ -138,6 +136,5 @@ class FeatureCoSTEER(CoSTEER):
             evolving_version=2,
             scen=scen,
             max_loop=DS_RD_SETTING.coder_max_loop,
-            max_seconds=scen.real_debug_timeout() * settings.max_seconds_multiplier,
             **kwargs,
         )

rdagent/components/coder/data_science/model/__init__.py

@@ -1,8 +1,6 @@
 from pathlib import Path
-from typing import Dict
 
 from rdagent.app.data_science.conf import DS_RD_SETTING
-from rdagent.components.coder.CoSTEER import CoSTEER
 from rdagent.components.coder.CoSTEER.evaluators import (
     CoSTEERMultiEvaluator,
     CoSTEERSingleFeedback,
@@ -18,6 +16,7 @@ from rdagent.components.coder.data_science.model.eval import (
     ModelGeneralCaseSpecEvaluator,
 )
 from rdagent.components.coder.data_science.model.exp import ModelTask
+from rdagent.components.coder.data_science.share.ds_costeer import DSCoSTEER
 from rdagent.core.exception import CoderError
 from rdagent.core.experiment import FBWorkspace
 from rdagent.core.scenario import Scenario
@@ -148,7 +147,7 @@ class ModelMultiProcessEvolvingStrategy(MultiProcessEvolvingStrategy):
         return evo
 
 
-class ModelCoSTEER(CoSTEER):
+class ModelCoSTEER(DSCoSTEER):
     def __init__(
         self,
         scen: Scenario,
@@ -170,6 +169,5 @@ class ModelCoSTEER(CoSTEER):
             evolving_version=2,
             scen=scen,
             max_loop=DS_RD_SETTING.coder_max_loop,
-            max_seconds=scen.real_debug_timeout() * settings.max_seconds_multiplier,
             **kwargs,
         )

rdagent/components/coder/data_science/pipeline/__init__.py

@@ -25,7 +25,6 @@ File structure
 from pathlib import Path
 
 from rdagent.app.data_science.conf import DS_RD_SETTING
-from rdagent.components.coder.CoSTEER import CoSTEER
 from rdagent.components.coder.CoSTEER.evaluators import (
     CoSTEERMultiEvaluator,
     CoSTEERSingleFeedback,
@@ -39,6 +38,7 @@ from rdagent.components.coder.CoSTEER.knowledge_management import (
 from rdagent.components.coder.data_science.conf import DSCoderCoSTEERSettings
 from rdagent.components.coder.data_science.pipeline.eval import PipelineCoSTEEREvaluator
 from rdagent.components.coder.data_science.pipeline.exp import PipelineTask
+from rdagent.components.coder.data_science.share.ds_costeer import DSCoSTEER
 from rdagent.components.coder.data_science.share.eval import ModelDumpEvaluator
 from rdagent.core.exception import CoderError
 from rdagent.core.experiment import FBWorkspace
@@ -130,7 +130,7 @@ class PipelineMultiProcessEvolvingStrategy(MultiProcessEvolvingStrategy):
         return evo
 
 
-class PipelineCoSTEER(CoSTEER):
+class PipelineCoSTEER(DSCoSTEER):
     def __init__(
         self,
         scen: Scenario,
@@ -159,6 +159,5 @@ class PipelineCoSTEER(CoSTEER):
             evolving_version=2,
             scen=scen,
             max_loop=DS_RD_SETTING.coder_max_loop,
-            max_seconds=scen.real_debug_timeout() * settings.max_seconds_multiplier,
             **kwargs,
         )

rdagent/components/coder/data_science/raw_data_loader/__init__.py

@@ -22,13 +22,10 @@ File structure
     - Each coder could be tested.
 """
 
-import json
 import re
 from pathlib import Path
-from typing import Dict
 
 from rdagent.app.data_science.conf import DS_RD_SETTING
-from rdagent.components.coder.CoSTEER import CoSTEER
 from rdagent.components.coder.CoSTEER.evaluators import (
     CoSTEERMultiEvaluator,
     CoSTEERSingleFeedback,
@@ -47,6 +44,7 @@ from rdagent.components.coder.data_science.raw_data_loader.eval import (
     DataLoaderCoSTEEREvaluator,
 )
 from rdagent.components.coder.data_science.raw_data_loader.exp import DataLoaderTask
+from rdagent.components.coder.data_science.share.ds_costeer import DSCoSTEER
 from rdagent.core.exception import CoderError
 from rdagent.core.experiment import FBWorkspace
 from rdagent.core.scenario import Scenario
@@ -197,7 +195,7 @@ class DataLoaderMultiProcessEvolvingStrategy(MultiProcessEvolvingStrategy):
         return evo
 
 
-class DataLoaderCoSTEER(CoSTEER):
+class DataLoaderCoSTEER(DSCoSTEER):
     def __init__(
         self,
         scen: Scenario,
@@ -218,7 +216,6 @@ class DataLoaderCoSTEER(CoSTEER):
             evolving_version=2,
             scen=scen,
             max_loop=DS_RD_SETTING.coder_max_loop,
-            max_seconds=scen.real_debug_timeout() * settings.max_seconds_multiplier,
             **kwargs,
         )
 

rdagent/components/coder/data_science/share/ds_costeer.py

@@ -0,0 +1,9 @@
+from rdagent.components.coder.CoSTEER import CoSTEER
+
+
+class DSCoSTEER(CoSTEER):
+    def get_develop_max_seconds(self) -> int | None:
+        """
+        The coder uses the scenario's real debug timeout as the maximum seconds for development.
+        """
+        return int(self.scen.real_debug_timeout() * self.settings.max_seconds_multiplier)

rdagent/components/coder/data_science/workflow/__init__.py

@@ -1,8 +1,4 @@
-import json
-from typing import Dict
-
 from rdagent.app.data_science.conf import DS_RD_SETTING
-from rdagent.components.coder.CoSTEER import CoSTEER
 from rdagent.components.coder.CoSTEER.evaluators import (
     CoSTEERMultiEvaluator,
     CoSTEERSingleFeedback,
@@ -14,6 +10,7 @@ from rdagent.components.coder.CoSTEER.knowledge_management import (
     CoSTEERQueriedKnowledge,
 )
 from rdagent.components.coder.data_science.conf import DSCoderCoSTEERSettings
+from rdagent.components.coder.data_science.share.ds_costeer import DSCoSTEER
 from rdagent.components.coder.data_science.workflow.eval import (
     WorkflowGeneralCaseSpecEvaluator,
 )
@@ -111,7 +108,7 @@ class WorkflowMultiProcessEvolvingStrategy(MultiProcessEvolvingStrategy):
         return evo
 
 
-class WorkflowCoSTEER(CoSTEER):
+class WorkflowCoSTEER(DSCoSTEER):
     def __init__(
         self,
         scen: Scenario,
@@ -131,6 +128,5 @@ class WorkflowCoSTEER(CoSTEER):
             evolving_version=2,
             scen=scen,
             max_loop=DS_RD_SETTING.coder_max_loop,
-            max_seconds=scen.real_debug_timeout() * settings.max_seconds_multiplier,
             **kwargs,
         )