dragon-ml-toolbox 19.12.0__tar.gz → 19.12.1__tar.gz

{dragon_ml_toolbox-19.12.0/dragon_ml_toolbox.egg-info → dragon_ml_toolbox-19.12.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: dragon-ml-toolbox
-Version: 19.12.0
+Version: 19.12.1
 Summary: Complete pipelines and helper tools for data science and machine learning projects.
 Author-email: Karl Luigi Loza Vidaurre <luigiloza@gmail.com>
 License-Expression: MIT

{dragon_ml_toolbox-19.12.0 → dragon_ml_toolbox-19.12.1/dragon_ml_toolbox.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: dragon-ml-toolbox
-Version: 19.12.0
+Version: 19.12.1
 Summary: Complete pipelines and helper tools for data science and machine learning projects.
 Author-email: Karl Luigi Loza Vidaurre <luigiloza@gmail.com>
 License-Expression: MIT

{dragon_ml_toolbox-19.12.0 → dragon_ml_toolbox-19.12.1}/ml_tools/_core/_IO_tools.py

@@ -434,8 +434,8 @@ def train_logger(train_config: Union[dict, Any],
     Logs training data to JSON, adding a timestamp to the filename.
     
     Args:
-        train_config (dict | Any): Training configuration parameters.
-        model_parameters (dict | Any): Model parameters.
+        train_config (dict | Any): Training configuration parameters. If object, must have a `.to_log()` method returning a dict.
+        model_parameters (dict | Any): Model parameters. If object, must have a `.to_log()` method returning a dict.
         train_history (dict | None): Training history log.
         save_directory (str | Path): Directory to save the log file.
     """
@@ -443,6 +443,9 @@ def train_logger(train_config: Union[dict, Any],
     if not isinstance(train_config, dict):
         if hasattr(train_config, "to_log") and callable(getattr(train_config, "to_log")):
             train_config_dict: dict = train_config.to_log()
+            if not isinstance(train_config_dict, dict):
+                _LOGGER.error("'train_config.to_log()' did not return a dictionary.")
+                raise ValueError()
         else:
             _LOGGER.error("'train_config' must be a dict or an object with a 'to_log()' method.")
             raise ValueError()
@@ -458,6 +461,9 @@ def train_logger(train_config: Union[dict, Any],
     if not isinstance(model_parameters, dict):
         if hasattr(model_parameters, "to_log") and callable(getattr(model_parameters, "to_log")):
             model_parameters_dict: dict = model_parameters.to_log()
+            if not isinstance(model_parameters_dict, dict):
+                _LOGGER.error("'model_parameters.to_log()' did not return a dictionary.")
+                raise ValueError()
         else:
             _LOGGER.error("'model_parameters' must be a dict or an object with a 'to_log()' method.")
             raise ValueError()

{dragon_ml_toolbox-19.12.0 → dragon_ml_toolbox-19.12.1}/ml_tools/_core/_ML_configuration.py

@@ -660,18 +660,27 @@ class DragonTrainingConfig(_BaseModelParams):
                  initial_learning_rate: float,
                  batch_size: int,
                  random_state: int = 101,
-                 early_stop_patience: Optional[int] = None,
-                 scheduler_patience: Optional[int] = None,
-                 scheduler_lr_factor: Optional[float] = None,
+                #  early_stop_patience: Optional[int] = None,
+                #  scheduler_patience: Optional[int] = None,
+                #  scheduler_lr_factor: Optional[float] = None,
                  **kwargs: Any) -> None:
+        """  
+        Args:
+            validation_size (float): Proportion of data for validation set.
+            test_size (float): Proportion of data for test set.
+            initial_learning_rate (float): Starting learning rate.
+            batch_size (int): Number of samples per training batch.
+            random_state (int): Seed for reproducibility.
+            **kwargs: Additional training parameters as key-value pairs.
+        """
         self.validation_size = validation_size
         self.test_size = test_size
         self.initial_learning_rate = initial_learning_rate
         self.batch_size = batch_size
         self.random_state = random_state
-        self.early_stop_patience = early_stop_patience
-        self.scheduler_patience = scheduler_patience
-        self.scheduler_lr_factor = scheduler_lr_factor
+        # self.early_stop_patience = early_stop_patience
+        # self.scheduler_patience = scheduler_patience
+        # self.scheduler_lr_factor = scheduler_lr_factor
         
         # Process kwargs with validation
         for key, value in kwargs.items():

{dragon_ml_toolbox-19.12.0 → dragon_ml_toolbox-19.12.1}/ml_tools/_core/_ML_finalize_handler.py

@@ -51,7 +51,7 @@ class FinalizedFileHandler:
         self._initial_sequence: Optional[np.ndarray] = None
         self._target_name: Optional[str] = None
         self._target_names: Optional[list[str]] = None
-        self._model_state_dict: Optional[Any] = None
+        self._model_state_dict: Optional[dict[str, Any]] = None
         
         # Set warning outputs
         self._verbose: bool=True
@@ -90,7 +90,7 @@ class FinalizedFileHandler:
             
         else:
             # It is a dict, but missing the keys, assume it is the raw state dict
-            _LOGGER.info(f"File '{pth_path.name}' does not have the required keys for a finalized-file. Treating it as raw PyTorch state dictionary.")
+            _LOGGER.warning(f"File '{pth_path.name}' does not have the required keys for a Dragon-ML finalized-file. Keys found:\n    {list(pth_file_content.keys())}")
             self._model_state_dict = pth_file_content
     
             
@@ -113,9 +113,10 @@ class FinalizedFileHandler:
         return self._task
     
     @property
-    def model_state_dict(self):
+    def model_state_dict(self) -> dict[str, Any]:
         """Returns the model state dictionary."""
-        return self._model_state_dict
+        # No need to check for None, as it is guaranteed to be set in __init__
+        return self._model_state_dict # type: ignore
     
     @property
     def epoch(self) -> Optional[int]:

{dragon_ml_toolbox-19.12.0 → dragon_ml_toolbox-19.12.1}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "dragon-ml-toolbox"
-version = "19.12.0"
+version = "19.12.1"
 description = "Complete pipelines and helper tools for data science and machine learning projects."
 authors = [
     { name = "Karl Luigi Loza Vidaurre", email = "luigiloza@gmail.com" }