autogluon.tabular 1.2.1b20250417__tar.gz → 1.2.1b20250418__tar.gz

{autogluon.tabular-1.2.1b20250417 → autogluon.tabular-1.2.1b20250418}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: autogluon.tabular
-Version: 1.2.1b20250417
+Version: 1.2.1b20250418
 Summary: Fast and Accurate ML in 3 Lines of Code
 Home-page: https://github.com/autogluon/autogluon
 Author: AutoGluon Community

{autogluon.tabular-1.2.1b20250417 → autogluon.tabular-1.2.1b20250418}/src/autogluon/tabular/configs/config_helper.py

@@ -9,7 +9,7 @@ from autogluon.core.scheduler import scheduler_factory
 from autogluon.features import AutoMLPipelineFeatureGenerator
 from autogluon.tabular.configs.hyperparameter_configs import hyperparameter_config_dict
 from autogluon.tabular.configs.presets_configs import tabular_presets_dict
-from autogluon.tabular.trainer.model_presets.presets import MODEL_TYPES
+from autogluon.tabular.register import ag_model_register
 
 
 class FeatureGeneratorBuilder:
@@ -107,6 +107,10 @@ class ConfigBuilder:
     def __init__(self):
         self.config = {}
 
+    def _valid_keys(self):
+        valid_keys = [m for m in ag_model_register.keys if m not in ["ENS_WEIGHTED", "SIMPLE_ENS_WEIGHTED"]]
+        return valid_keys
+
     def presets(self, presets: Union[str, list, dict]) -> ConfigBuilder:
         """
         List of preset configurations for various arguments in `fit()`. Can significantly impact predictive accuracy, memory-footprint, and inference latency of trained models, and various other properties of the returned `predictor`.
@@ -137,7 +141,7 @@ class ConfigBuilder:
         return self
 
     def hyperparameters(self, hyperparameters: Union[str, dict]) -> ConfigBuilder:
-        valid_keys = [m for m in MODEL_TYPES.keys() if m not in ["ENS_WEIGHTED", "SIMPLE_ENS_WEIGHTED"]]
+        valid_keys = self._valid_keys()
         valid_str_values = list(hyperparameter_config_dict.keys())
         if isinstance(hyperparameters, str):
             assert hyperparameters in hyperparameter_config_dict, f"{hyperparameters} is not one of the valid presets {valid_str_values}"
@@ -270,7 +274,7 @@ class ConfigBuilder:
         Useful when a particular model type such as 'KNN' or 'custom' is not desired but altering the `hyperparameters` dictionary is difficult or time-consuming.
             Example: To exclude both 'KNN' and 'custom' models, specify `excluded_model_types=['KNN', 'custom']`.
         """
-        valid_keys = [m for m in MODEL_TYPES.keys() if m not in ["ENS_WEIGHTED", "SIMPLE_ENS_WEIGHTED"]]
+        valid_keys = self._valid_keys()
         if not isinstance(models, list):
             models = [models]
         for model in models:
@@ -285,7 +289,7 @@ class ConfigBuilder:
         Useful when only the particular models should be trained such as 'KNN' or 'custom', but altering the `hyperparameters` dictionary is difficult or time-consuming.
             Example: To keep only 'KNN' and 'custom' models, specify `included_model_types=['KNN', 'custom']`.
         """
-        valid_keys = [m for m in MODEL_TYPES.keys() if m not in ["ENS_WEIGHTED", "SIMPLE_ENS_WEIGHTED"]]
+        valid_keys = self._valid_keys()
         if not isinstance(models, list):
             models = [models]
 

{autogluon.tabular-1.2.1b20250417 → autogluon.tabular-1.2.1b20250418}/src/autogluon/tabular/predictor/predictor.py

@@ -55,7 +55,7 @@ from ..configs.hyperparameter_configs import get_hyperparameter_config
 from ..configs.presets_configs import tabular_presets_alias, tabular_presets_dict
 from ..learner import AbstractTabularLearner, DefaultLearner
 from ..trainer.abstract_trainer import AbstractTabularTrainer
-from ..trainer.model_presets.presets import MODEL_TYPES
+from ..register import ag_model_register
 from ..version import __version__
 
 logger = logging.getLogger(__name__)  # return autogluon root logger
@@ -516,7 +516,6 @@ class TabularPredictor:
                     'FT_TRANSFORMER' (Tabular Transformer, GPU is recommended. Does not scale well to >100 features.)
                     'FASTTEXT' (FastText. Note: Has not been tested for a long time.)
                     'TABPFN' (TabPFN. Does not scale well to >100 features or >1000 rows, and does not support regression. Extremely slow inference speed.)
-                    'VW' (VowpalWabbit. Note: Has not been tested for a long time.)
                     'AG_TEXT_NN' (Multimodal Text+Tabular model, GPU is required. Recommended to instead use its successor, 'AG_AUTOMM'.)
                     'AG_IMAGE_NN' (Image model, GPU is required. Recommended to instead use its successor, 'AG_AUTOMM'.)
                 If a certain key is missing from hyperparameters, then `fit()` will not train any models of that type. Omitting a model key from hyperparameters is equivalent to including this model key in `excluded_model_types`.
@@ -977,17 +976,10 @@ class TabularPredictor:
                         sklearn CountVectorizer object to use in TextNgramFeatureGenerator.
                         Only used if `enable_text_ngram_features=True`.
             unlabeled_data : pd.DataFrame, default = None
-                [Experimental Parameter]
-                Collection of data without labels that we can use to pretrain on. This is the same schema as train_data, except
-                without the labels. Currently, unlabeled_data is only used for pretraining a TabTransformer model.
-                If you do not specify 'TRANSF' with unlabeled_data, then no pretraining will occur and unlabeled_data will be ignored!
-                After the pretraining step, we will finetune using the TabTransformer model as well. If TabTransformer is ensembled
-                with other models, like in typical AutoGluon fashion, then the output of this "pretrain/finetune" will be ensembled
-                with other models, which will not used the unlabeled_data. The "pretrain/finetune flow" is also known as semi-supervised learning.
-                The typical use case for unlabeled_data is to add signal to your model where you may not have sufficient training
-                data. e.g. 500 hand-labeled samples (perhaps a hard human task), whole data set (unlabeled) is thousands/millions.
-                However, this isn't the only use case. Given enough unlabeled data(millions of rows), you may see improvements
-                to any amount of labeled data.
+                [Experimental Parameter] UNUSED.
+                Collection of data without labels that we can use to pretrain on.
+                This is the same schema as train_data, except without the labels.
+                Currently, unlabeled_data is not used by any model.
             verbosity : int
                 If specified, overrides the existing `predictor.verbosity` value.
             raise_on_model_failure: bool, default = False
@@ -5680,12 +5672,13 @@ class TabularPredictor:
             for key in hyperparameters:
                 models_in_hyperparameters.add(key)
         models_in_hyperparameters_raw_text_compatible = []
+        model_key_to_cls_map = ag_model_register.key_to_cls_map()
         for m in models_in_hyperparameters:
             if isinstance(m, str):
                 # TODO: Technically the use of MODEL_TYPES here is a hack since we should derive valid types from trainer,
                 #  but this is required prior to trainer existing.
-                if m in MODEL_TYPES:
-                    m = MODEL_TYPES[m]
+                if m in model_key_to_cls_map:
+                    m = model_key_to_cls_map[m]
                 else:
                     continue
             if m._get_class_tags().get("handles_text", False):

{autogluon.tabular-1.2.1b20250417 → autogluon.tabular-1.2.1b20250418}/src/autogluon/tabular/trainer/auto_trainer.py

@@ -5,8 +5,9 @@ from autogluon.core.utils import generate_train_test_split
 
 from ..models.lgb.lgb_model import LGBModel
 from .abstract_trainer import AbstractTabularTrainer
-from .model_presets.presets import MODEL_TYPES, get_preset_models
+from .model_presets.presets import get_preset_models
 from .model_presets.presets_distill import get_preset_models_distillation
+from ..register import ag_model_register
 
 logger = logging.getLogger(__name__)
 
@@ -186,4 +187,4 @@ class AutoTrainer(AbstractTabularTrainer):
         return super().compile(model_names=model_names, with_ancestors=with_ancestors, compiler_configs=compiler_configs)
 
     def _get_model_types_map(self) -> dict[str, AbstractModel]:
-        return MODEL_TYPES
+        return ag_model_register.key_to_cls_map()

{autogluon.tabular-1.2.1b20250417 → autogluon.tabular-1.2.1b20250418}/src/autogluon/tabular/trainer/model_presets/presets.py

@@ -33,31 +33,8 @@ from ...version import __version__
 
 logger = logging.getLogger(__name__)
 
-# TODO: Replace with ag_model_register
-# Problem type specific model priority overrides (will update default values in DEFAULT_MODEL_PRIORITY)
-PROBLEM_TYPE_MODEL_PRIORITY = {
-    MULTICLASS: dict(
-        FASTAI=95,
-    ),
-}
-
-# TODO: Replace with ag_model_register
-DEFAULT_SOFTCLASS_PRIORITY = dict(
-    GBM=100,
-    RF=80,
-    CAT=60,
-    custom=0,
-)
-
 DEFAULT_CUSTOM_MODEL_PRIORITY = 0
 
-# FIXME: Don't do this, use ag_model_register lazily so users can register custom models before calling fit
-DEFAULT_MODEL_PRIORITY = {ag_model_register.key(model_cls): ag_model_register.priority(model_cls) for model_cls in ag_model_register.model_cls_list}
-DEFAULT_MODEL_NAMES = ag_model_register.name_map()
-REGISTERED_MODEL_CLS_LST = ag_model_register.model_cls_list
-MODEL_TYPES = ag_model_register.key_to_cls_map()
-
-
 VALID_AG_ARGS_KEYS = {
     "name",
     "name_main",
@@ -122,9 +99,10 @@ def get_preset_models(
         invalid_name_set.update(invalid_model_names)
 
     if default_priorities is None:
-        default_priorities = copy.deepcopy(DEFAULT_MODEL_PRIORITY)
-        if problem_type in PROBLEM_TYPE_MODEL_PRIORITY:
-            default_priorities.update(PROBLEM_TYPE_MODEL_PRIORITY[problem_type])
+        priority_cls_map = ag_model_register.priority_map(problem_type=problem_type)
+        default_priorities = {
+            ag_model_register.key(model_cls): priority for model_cls, priority in priority_cls_map.items()
+        }
 
     level_key = level if level in hyperparameters.keys() else "default"
     if level_key not in hyperparameters.keys() and level_key == "default":
@@ -197,20 +175,22 @@ def clean_model_cfg(model_cfg: dict, model_type=None, ag_args=None, ag_args_ense
     if model_cfg[AG_ARGS]["model_type"] is None:
         raise AssertionError(f"model_type was not specified for model! Model: {model_cfg}")
     model_type = model_cfg[AG_ARGS]["model_type"]
+    model_types = ag_model_register.key_to_cls_map()
     if not inspect.isclass(model_type):
-        if model_type not in MODEL_TYPES:
-            raise AssertionError(f"Unknown model type specified in hyperparameters: '{model_type}'. Valid model types: {list(MODEL_TYPES.keys())}")
-        model_type = MODEL_TYPES[model_type]
+        if model_type not in model_types:
+            raise AssertionError(f"Unknown model type specified in hyperparameters: '{model_type}'. Valid model types: {list(model_types.keys())}")
+        model_type = model_types[model_type]
     elif not issubclass(model_type, AbstractModel):
         logger.warning(
             f"Warning: Custom model type {model_type} does not inherit from {AbstractModel}. This may lead to instability. Consider wrapping {model_type} with an implementation of {AbstractModel}!"
         )
     else:
-        logger.log(20, f"Custom Model Type Detected: {model_type}")
+        if not ag_model_register.exists(model_type):
+            logger.log(20, f"Custom Model Type Detected: {model_type}")
     model_cfg[AG_ARGS]["model_type"] = model_type
     model_type_real = model_cfg[AG_ARGS]["model_type"]
     if not inspect.isclass(model_type_real):
-        model_type_real = MODEL_TYPES[model_type_real]
+        model_type_real = model_types[model_type_real]
     default_ag_args = model_type_real._get_default_ag_args()
     if ag_args is not None:
         model_extra_ag_args = ag_args.copy()
@@ -300,10 +280,13 @@ def model_factory(
         invalid_name_set = set()
     model_type = model[AG_ARGS]["model_type"]
     if not inspect.isclass(model_type):
-        model_type = MODEL_TYPES[model_type]
+        model_type = ag_model_register.key_to_cls(model_type)
     name_orig = model[AG_ARGS].get("name", None)
     if name_orig is None:
-        name_main = model[AG_ARGS].get("name_main", DEFAULT_MODEL_NAMES.get(model_type, model_type.__name__))
+        ag_name = model_type.ag_name
+        if ag_name is None:
+            ag_name = model_type.__name__
+        name_main = model[AG_ARGS].get("name_main", ag_name)
         name_prefix = model[AG_ARGS].get("name_prefix", "")
         name_suff = model[AG_ARGS].get("name_suffix", "")
         name_orig = name_prefix + name_main + name_suff
@@ -372,7 +355,6 @@ def get_preset_models_softclass(hyperparameters, invalid_model_names: list = Non
         problem_type=SOFTCLASS,
         eval_metric=soft_log_loss,
         hyperparameters=hyperparameters_standard,
-        default_priorities=DEFAULT_SOFTCLASS_PRIORITY,
         invalid_model_names=invalid_model_names,
         **kwargs,
     )

{autogluon.tabular-1.2.1b20250417 → autogluon.tabular-1.2.1b20250418}/src/autogluon/tabular/version.py

@@ -1,4 +1,4 @@
 """This is the autogluon version file."""
 
-__version__ = "1.2.1b20250417"
+__version__ = "1.2.1b20250418"
 __lite__ = False

{autogluon.tabular-1.2.1b20250417 → autogluon.tabular-1.2.1b20250418}/src/autogluon.tabular.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: autogluon.tabular
-Version: 1.2.1b20250417
+Version: 1.2.1b20250418
 Summary: Fast and Accurate ML in 3 Lines of Code
 Home-page: https://github.com/autogluon/autogluon
 Author: AutoGluon Community

{autogluon.tabular-1.2.1b20250417 → autogluon.tabular-1.2.1b20250418}/src/autogluon.tabular.egg-info/requires.txt

@@ -3,20 +3,20 @@ scipy<1.16,>=1.5.4
 pandas<2.3.0,>=2.0.0
 scikit-learn<1.7.0,>=1.4.0
 networkx<4,>=3.0
-autogluon.core==1.2.1b20250417
-autogluon.features==1.2.1b20250417
+autogluon.core==1.2.1b20250418
+autogluon.features==1.2.1b20250418
 
 [all]
-xgboost<3.1,>=2.0
-numpy<2.0.0,>=1.25
 einops<0.9,>=0.7
+catboost<1.3,>=1.2
+xgboost<3.1,>=2.0
 torch<2.7,>=2.2
-autogluon.core[all]==1.2.1b20250417
-fastai<2.9,>=2.3.1
+autogluon.core[all]==1.2.1b20250418
 huggingface_hub[torch]
-catboost<1.3,>=1.2
-lightgbm<4.7,>=4.0
 spacy<3.8
+fastai<2.9,>=2.3.1
+numpy<2.0.0,>=1.25
+lightgbm<4.7,>=4.0
 
 [catboost]
 numpy<2.0.0,>=1.25
@@ -34,7 +34,7 @@ imodels<2.1.0,>=1.3.10
 lightgbm<4.7,>=4.0
 
 [ray]
-autogluon.core[all]==1.2.1b20250417
+autogluon.core[all]==1.2.1b20250418
 
 [skex]
 scikit-learn-intelex<2025.5,>=2024.0