Add FE enabled flag (#127)

autogluon · Nov 14, 2024 · 7559e8d · 7559e8d
1 parent c0a1179
commit 7559e8d
Show file tree

Hide file tree

Showing 6 changed files with 45 additions and 42 deletions.
diff --git a/README.md b/README.md
@@ -109,9 +109,9 @@ the `config_overrides` parameter with Hydra syntax from the command line.
 Here’s an example command with some configuration overrides:
 
 ```bash
-aga run toy_data --config_overrides "feature_transformers=[], autogluon.predictor_fit_kwargs.time_limit=3600"
+aga run toy_data --config_overrides "feature_transformers.enabled=False, autogluon.predictor_fit_kwargs.time_limit=3600"
 
 # OR
 
-aga run toy_data --config_overrides "feature_transformers=[]" --config_overrides "autogluon.predictor_fit_kwargs.time_limit=3600"
+aga run toy_data --config_overrides "feature_transformers.enabled=False" --config_overrides "autogluon.predictor_fit_kwargs.time_limit=3600"
 ```
diff --git a/src/autogluon_assistant/assistant.py b/src/autogluon_assistant/assistant.py
@@ -47,8 +47,8 @@ def __init__(self, config: DictConfig) -> None:
         self.config = config
         self.llm: Union[AssistantChatOpenAI, AssistantChatBedrock] = LLMFactory.get_chat_model(config.llm)
         self.predictor = AutogluonTabularPredictor(config.autogluon)
-        self.feature_transformers_config = config.feature_transformers
-        self.use_feature_transformers = config.use_feature_transformers
+        self.feature_transformers_config = config.feature_transformers.transformers
+        self.use_feature_transformers = config.feature_transformers.enabled
 
     def describe(self) -> Dict[str, Any]:
         return {

diff --git a/src/autogluon_assistant/configs/best_quality.yaml b/src/autogluon_assistant/configs/best_quality.yaml
@@ -5,19 +5,20 @@ save_artifacts:
   enabled: False
   append_timestamp: True
   path: "./aga-artifacts"
-use_feature_transformers: True
 feature_transformers:
-  - _target_: autogluon_assistant.transformer.CAAFETransformer
-    eval_model: lightgbm
-    llm_provider: ${llm.provider}
-    llm_model: ${llm.model}
-    num_iterations: 5
-    optimization_metric: roc
-  - _target_: autogluon_assistant.transformer.OpenFETransformer
-    n_jobs: 1
-    num_features_to_keep: 10
-  - _target_: autogluon_assistant.transformer.PretrainedEmbeddingTransformer
-    model_name: 'all-mpnet-base-v2'
+  enabled: True
+  transformers:
+    - _target_: autogluon_assistant.transformer.CAAFETransformer
+      eval_model: lightgbm
+      llm_provider: ${llm.provider}
+      llm_model: ${llm.model}
+      num_iterations: 5
+      optimization_metric: roc
+    - _target_: autogluon_assistant.transformer.OpenFETransformer
+      n_jobs: 1
+      num_features_to_keep: 10
+    - _target_: autogluon_assistant.transformer.PretrainedEmbeddingTransformer
+      model_name: 'all-mpnet-base-v2'
 autogluon:
   predictor_init_kwargs: {}
   predictor_fit_kwargs:

diff --git a/src/autogluon_assistant/configs/high_quality.yaml b/src/autogluon_assistant/configs/high_quality.yaml
@@ -5,19 +5,20 @@ save_artifacts:
   enabled: False
   append_timestamp: True
   path: "./aga-artifacts"
-use_feature_transformers: False
 feature_transformers:
-  - _target_: autogluon_assistant.transformer.CAAFETransformer
-    eval_model: lightgbm
-    llm_provider: ${llm.provider}
-    llm_model: ${llm.model}
-    num_iterations: 5
-    optimization_metric: roc
-  - _target_: autogluon_assistant.transformer.OpenFETransformer
-    n_jobs: 1
-    num_features_to_keep: 10
-  - _target_: autogluon_assistant.transformer.PretrainedEmbeddingTransformer
-    model_name: 'all-mpnet-base-v2'
+  enabled: False
+  transformers:
+    - _target_: autogluon_assistant.transformer.CAAFETransformer
+      eval_model: lightgbm
+      llm_provider: ${llm.provider}
+      llm_model: ${llm.model}
+      num_iterations: 5
+      optimization_metric: roc
+    - _target_: autogluon_assistant.transformer.OpenFETransformer
+      n_jobs: 1
+      num_features_to_keep: 10
+    - _target_: autogluon_assistant.transformer.PretrainedEmbeddingTransformer
+      model_name: 'all-mpnet-base-v2'
 autogluon:
   predictor_init_kwargs: {}
   predictor_fit_kwargs:

diff --git a/src/autogluon_assistant/configs/medium_quality.yaml b/src/autogluon_assistant/configs/medium_quality.yaml
@@ -5,19 +5,20 @@ save_artifacts:
   enabled: False
   append_timestamp: True
   path: "./aga-artifacts"
-use_feature_transformers: False
 feature_transformers:
-  - _target_: autogluon_assistant.transformer.CAAFETransformer
-    eval_model: lightgbm
-    llm_provider: ${llm.provider}
-    llm_model: ${llm.model}
-    num_iterations: 5
-    optimization_metric: roc
-  - _target_: autogluon_assistant.transformer.OpenFETransformer
-    n_jobs: 1
-    num_features_to_keep: 10
-  - _target_: autogluon_assistant.transformer.PretrainedEmbeddingTransformer
-    model_name: 'all-mpnet-base-v2'
+  enabled: False
+  transformers:
+    - _target_: autogluon_assistant.transformer.CAAFETransformer
+      eval_model: lightgbm
+      llm_provider: ${llm.provider}
+      llm_model: ${llm.model}
+      num_iterations: 5
+      optimization_metric: roc
+    - _target_: autogluon_assistant.transformer.OpenFETransformer
+      n_jobs: 1
+      num_features_to_keep: 10
+    - _target_: autogluon_assistant.transformer.PretrainedEmbeddingTransformer
+      model_name: 'all-mpnet-base-v2'
 autogluon:
   predictor_init_kwargs: {}
   predictor_fit_kwargs:

diff --git a/src/autogluon_assistant/ui/pages/task.py b/src/autogluon_assistant/ui/pages/task.py
@@ -51,9 +51,9 @@ def update_config_overrides():
         config_overrides.append(f"llm.provider={PROVIDER_MAPPING[st.session_state.llm]}")
 
     if st.session_state.feature_generation:
-        config_overrides.append("use_feature_transformers=True")
+        config_overrides.append("feature_transformers.enabled=True")
     else:
-        config_overrides.append("use_feature_transformers=False")
+        config_overrides.append("feature_transformers.enabled=False")
 
     st.session_state.config_overrides = config_overrides