torch-molecule 0.1.0.dev2__tar.gz → 0.1.1__tar.gz

{torch_molecule-0.1.0.dev2 → torch_molecule-0.1.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: torch_molecule
-Version: 0.1.0.dev2
+Version: 0.1.1
 Summary: Deep learning packages for molecular discovery with a simple sklearn-style interface
 Author-email: Gang Liu <gliu7@nd.edu>
 License: MIT
@@ -16,12 +16,9 @@ License-File: LICENSE
 Requires-Dist: torch>=2.2.0
 Requires-Dist: torch-geometric>=2.6.1
 Requires-Dist: numpy
-Requires-Dist: pandas>=2.2.3
-Requires-Dist: click
 Requires-Dist: huggingface_hub>=0.22.2
 Requires-Dist: joblib>=1.3.2
 Requires-Dist: networkx>=3.2.1
-Requires-Dist: PyYAML>=6.0.2
 Requires-Dist: rdkit>=2023.9.5
 Requires-Dist: scikit_learn>=1.4.1.post1
 Requires-Dist: scipy>=1.14.1
@@ -55,8 +52,6 @@ Dynamic: license-file
 
 `torch-molecule` is a package under active development that facilitates molecular discovery through deep learning, featuring a user-friendly, `sklearn`-style interface. It includes model checkpoints for efficient deployment and benchmarking across a range of molecular tasks. Currently, the package focuses on three main components: **Predictive Models**, **Generative Models**, and **Representation Models**. See the [List of Supported Models](#list-of-supported-models) section for all available models.
 
-> **Note**: This project is in active development, and features may change.
-
 ## Installation
 
 1. **Create a Conda environment**:
@@ -65,26 +60,25 @@ Dynamic: license-file
    conda activate torch_molecule
    ```
 
-2. **Install `torch_molecule` from GitHub**:
+2. **Install using pip (0.1.0)**:
 
-   Clone the repository:
    ```bash
-   git clone https://github.com/liugangcode/torch-molecule
+   pip install torch-molecule
    ```
 
-   Install the requirements:
-   ```bash
-   pip install -r requirements.txt
-   ```
+3. **Install from source for the latest version**:
+
+   Clone the repository:
 
-   Editable install:
    ```bash
-   pip install -e .
+   git clone https://github.com/liugangcode/torch-molecule
+   cd torch-molecule
    ```
 
-3. **Install `torch_molecule` from PyPI** (Legacy):
+   Install:
+
    ```bash
-   pip install -i https://test.pypi.org/simple/ torch-molecule
+   pip install .
    ```
 
 ### Additional Packages
@@ -95,6 +89,12 @@ Dynamic: license-file
 | BFGNNMolecularPredictor | torch-scatter |
 | GRINMolecularPredictor | torch-scatter |
 
+**For models that require `torch-scatter`: Install using the following command: `pip install torch-scatter -f https://data.pyg.org/whl/torch-${TORCH}+${CUDA}.html`, e.g.,
+
+> `pip install torch-scatter -f https://data.pyg.org/whl/torch-2.7.1+cu128.html`
+
+**For models that require `transformers`:** `pip install transformers`
+
 ## Usage
 
 Refer to the `tests` folder for more use cases.
@@ -236,8 +236,8 @@ See the structure of `torch_molecule` with the command `tree -L 2 torch_molecule
 2. **Generative Models**: Done: Graph DiT, GraphGA, DiGress, GDS, MolGPT TODO: more
 3. **Representation Models**: Done: MoAMa, AttrMasking, ContextPred, EdgePred. Many pretrained models from HF. TODO: checkpoints, more 
 
-## Acknowledgements
+> **Note**: This project is in active development, and features may change.
 
-This project is under active development, and some features may change over time.
+## Acknowledgements
 
 The project template was adapted from [https://github.com/lwaekfjlk/python-project-template](https://github.com/lwaekfjlk/python-project-template). We thank the authors for their contribution to the open-source community.

{torch_molecule-0.1.0.dev2 → torch_molecule-0.1.1}/README.md

@@ -19,8 +19,6 @@
 
 `torch-molecule` is a package under active development that facilitates molecular discovery through deep learning, featuring a user-friendly, `sklearn`-style interface. It includes model checkpoints for efficient deployment and benchmarking across a range of molecular tasks. Currently, the package focuses on three main components: **Predictive Models**, **Generative Models**, and **Representation Models**. See the [List of Supported Models](#list-of-supported-models) section for all available models.
 
-> **Note**: This project is in active development, and features may change.
-
 ## Installation
 
 1. **Create a Conda environment**:
@@ -29,26 +27,25 @@
    conda activate torch_molecule
    ```
 
-2. **Install `torch_molecule` from GitHub**:
+2. **Install using pip (0.1.0)**:
 
-   Clone the repository:
    ```bash
-   git clone https://github.com/liugangcode/torch-molecule
+   pip install torch-molecule
    ```
 
-   Install the requirements:
-   ```bash
-   pip install -r requirements.txt
-   ```
+3. **Install from source for the latest version**:
+
+   Clone the repository:
 
-   Editable install:
    ```bash
-   pip install -e .
+   git clone https://github.com/liugangcode/torch-molecule
+   cd torch-molecule
    ```
 
-3. **Install `torch_molecule` from PyPI** (Legacy):
+   Install:
+
    ```bash
-   pip install -i https://test.pypi.org/simple/ torch-molecule
+   pip install .
    ```
 
 ### Additional Packages
@@ -59,6 +56,12 @@
 | BFGNNMolecularPredictor | torch-scatter |
 | GRINMolecularPredictor | torch-scatter |
 
+**For models that require `torch-scatter`: Install using the following command: `pip install torch-scatter -f https://data.pyg.org/whl/torch-${TORCH}+${CUDA}.html`, e.g.,
+
+> `pip install torch-scatter -f https://data.pyg.org/whl/torch-2.7.1+cu128.html`
+
+**For models that require `transformers`:** `pip install transformers`
+
 ## Usage
 
 Refer to the `tests` folder for more use cases.
@@ -200,8 +203,8 @@ See the structure of `torch_molecule` with the command `tree -L 2 torch_molecule
 2. **Generative Models**: Done: Graph DiT, GraphGA, DiGress, GDS, MolGPT TODO: more
 3. **Representation Models**: Done: MoAMa, AttrMasking, ContextPred, EdgePred. Many pretrained models from HF. TODO: checkpoints, more 
 
-## Acknowledgements
+> **Note**: This project is in active development, and features may change.
 
-This project is under active development, and some features may change over time.
+## Acknowledgements
 
 The project template was adapted from [https://github.com/lwaekfjlk/python-project-template](https://github.com/lwaekfjlk/python-project-template). We thank the authors for their contribution to the open-source community.

{torch_molecule-0.1.0.dev2 → torch_molecule-0.1.1}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "torch_molecule"
-version = "0.1.0.dev2"
+version = "0.1.1"
 description = "Deep learning packages for molecular discovery with a simple sklearn-style interface"
 authors = [{name = "Gang Liu", email = "gliu7@nd.edu"}]
 readme = "README.md"
@@ -19,12 +19,9 @@ dependencies = [
     "torch>=2.2.0",
     "torch-geometric>=2.6.1",
     "numpy",
-    "pandas>=2.2.3",
-    "click",
     "huggingface_hub>=0.22.2",
     "joblib>=1.3.2",
     "networkx>=3.2.1",
-    "PyYAML>=6.0.2",
     "rdkit>=2023.9.5",
     "scikit_learn>=1.4.1.post1",
     "scipy>=1.14.1",

{torch_molecule-0.1.0.dev2 → torch_molecule-0.1.1}/torch_molecule/predictor/gnn/modeling_gnn.py

@@ -317,7 +317,7 @@ class GNNMolecularPredictor(BaseMolecularPredictor):
         y_val: Optional[Union[List, np.ndarray]] = None,
         X_unlbl: Optional[List[str]] = None,
         search_parameters: Optional[Dict[str, ParameterSpec]] = None,
-        n_trials: int = 10,
+        n_trials: int = 10, 
     ) -> "GNNMolecularPredictor":
         """Automatically find the best hyperparameters using Optuna optimization."""
         import optuna
@@ -565,19 +565,17 @@ class GNNMolecularPredictor(BaseMolecularPredictor):
                 best_eval = current_eval
                 best_state_dict = self.model.state_dict()
                 cnt_wait = 0
+                if self.verbose:
+                    print(
+                        f"Better result found at Epoch {epoch}: Loss = {np.mean(train_losses):.4f}, "
+                        f"{self.evaluate_name} = {best_eval:.4f}"
+                    )
             else:
                 cnt_wait += 1
                 if cnt_wait > self.patience:
                     if self.verbose:
                         print(f"Early stopping triggered after {epoch} epochs")
                     break
-            
-            if self.verbose and epoch % 10 == 0:
-                print(
-                    f"Epoch {epoch}: Loss = {np.mean(train_losses):.4f}, "
-                    f"{self.evaluate_name} = {current_eval:.4f}, "
-                    f"Best {self.evaluate_name} = {best_eval:.4f}"
-                )
 
         # Restore best model
         if best_state_dict is not None:
@@ -621,7 +619,11 @@ class GNNMolecularPredictor(BaseMolecularPredictor):
         self.model.eval()
         predictions = []
         with torch.no_grad():
-            for batch in tqdm(loader, disable=not self.verbose):
+            if self.verbose:
+                iterator = tqdm(loader, desc="Predicting")
+            else:
+                iterator = loader
+            for batch in iterator:
                 batch = batch.to(self.device)
                 out = self.model(batch)
                 predictions.append(out["prediction"].cpu().numpy())

{torch_molecule-0.1.0.dev2 → torch_molecule-0.1.1}/torch_molecule/predictor/smiles_transformer/modeling_transformer.py

@@ -240,7 +240,6 @@ class SMILESTransformerMolecularPredictor(LSTMMolecularPredictor):
                 mode='min' if not self.evaluate_higher_better else 'max',
                 factor=self.scheduler_factor,
                 patience=self.scheduler_patience,
-                verbose=self.verbose
             )
         
         return optimizer, scheduler

{torch_molecule-0.1.0.dev2 → torch_molecule-0.1.1}/torch_molecule.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: torch_molecule
-Version: 0.1.0.dev2
+Version: 0.1.1
 Summary: Deep learning packages for molecular discovery with a simple sklearn-style interface
 Author-email: Gang Liu <gliu7@nd.edu>
 License: MIT
@@ -16,12 +16,9 @@ License-File: LICENSE
 Requires-Dist: torch>=2.2.0
 Requires-Dist: torch-geometric>=2.6.1
 Requires-Dist: numpy
-Requires-Dist: pandas>=2.2.3
-Requires-Dist: click
 Requires-Dist: huggingface_hub>=0.22.2
 Requires-Dist: joblib>=1.3.2
 Requires-Dist: networkx>=3.2.1
-Requires-Dist: PyYAML>=6.0.2
 Requires-Dist: rdkit>=2023.9.5
 Requires-Dist: scikit_learn>=1.4.1.post1
 Requires-Dist: scipy>=1.14.1
@@ -55,8 +52,6 @@ Dynamic: license-file
 
 `torch-molecule` is a package under active development that facilitates molecular discovery through deep learning, featuring a user-friendly, `sklearn`-style interface. It includes model checkpoints for efficient deployment and benchmarking across a range of molecular tasks. Currently, the package focuses on three main components: **Predictive Models**, **Generative Models**, and **Representation Models**. See the [List of Supported Models](#list-of-supported-models) section for all available models.
 
-> **Note**: This project is in active development, and features may change.
-
 ## Installation
 
 1. **Create a Conda environment**:
@@ -65,26 +60,25 @@ Dynamic: license-file
    conda activate torch_molecule
    ```
 
-2. **Install `torch_molecule` from GitHub**:
+2. **Install using pip (0.1.0)**:
 
-   Clone the repository:
    ```bash
-   git clone https://github.com/liugangcode/torch-molecule
+   pip install torch-molecule
    ```
 
-   Install the requirements:
-   ```bash
-   pip install -r requirements.txt
-   ```
+3. **Install from source for the latest version**:
+
+   Clone the repository:
 
-   Editable install:
    ```bash
-   pip install -e .
+   git clone https://github.com/liugangcode/torch-molecule
+   cd torch-molecule
    ```
 
-3. **Install `torch_molecule` from PyPI** (Legacy):
+   Install:
+
    ```bash
-   pip install -i https://test.pypi.org/simple/ torch-molecule
+   pip install .
    ```
 
 ### Additional Packages
@@ -95,6 +89,12 @@ Dynamic: license-file
 | BFGNNMolecularPredictor | torch-scatter |
 | GRINMolecularPredictor | torch-scatter |
 
+**For models that require `torch-scatter`: Install using the following command: `pip install torch-scatter -f https://data.pyg.org/whl/torch-${TORCH}+${CUDA}.html`, e.g.,
+
+> `pip install torch-scatter -f https://data.pyg.org/whl/torch-2.7.1+cu128.html`
+
+**For models that require `transformers`:** `pip install transformers`
+
 ## Usage
 
 Refer to the `tests` folder for more use cases.
@@ -236,8 +236,8 @@ See the structure of `torch_molecule` with the command `tree -L 2 torch_molecule
 2. **Generative Models**: Done: Graph DiT, GraphGA, DiGress, GDS, MolGPT TODO: more
 3. **Representation Models**: Done: MoAMa, AttrMasking, ContextPred, EdgePred. Many pretrained models from HF. TODO: checkpoints, more 
 
-## Acknowledgements
+> **Note**: This project is in active development, and features may change.
 
-This project is under active development, and some features may change over time.
+## Acknowledgements
 
 The project template was adapted from [https://github.com/lwaekfjlk/python-project-template](https://github.com/lwaekfjlk/python-project-template). We thank the authors for their contribution to the open-source community.

{torch_molecule-0.1.0.dev2 → torch_molecule-0.1.1}/torch_molecule.egg-info/SOURCES.txt

@@ -1,7 +1,6 @@
 LICENSE
 README.md
 pyproject.toml
-tests/test_io.py
 torch_molecule/__init__.py
 torch_molecule.egg-info/PKG-INFO
 torch_molecule.egg-info/SOURCES.txt

{torch_molecule-0.1.0.dev2 → torch_molecule-0.1.1}/torch_molecule.egg-info/requires.txt

@@ -1,12 +1,9 @@
 torch>=2.2.0
 torch-geometric>=2.6.1
 numpy
-pandas>=2.2.3
-click
 huggingface_hub>=0.22.2
 joblib>=1.3.2
 networkx>=3.2.1
-PyYAML>=6.0.2
 rdkit>=2023.9.5
 scikit_learn>=1.4.1.post1
 scipy>=1.14.1

torch_molecule-0.1.0.dev2/tests/test_io.py

@@ -1,319 +0,0 @@
-import numpy as np
-import os
-import shutil
-from pathlib import Path
-from torch_molecule import (
-    GNNMolecularPredictor,
-    GREAMolecularPredictor,
-    SGIRMolecularPredictor,
-    SupervisedMolecularEncoder
-)
-from unittest.mock import patch, MagicMock
-import torch
-from huggingface_hub import HfApi
-import json
-
-# Test data
-SMILES_LIST = [
-    'CNC[C@H]1OCc2cnnn2CCCC(=O)N([C@H](C)CO)C[C@@H]1C',
-    'CNC[C@@H]1OCc2cnnn2CCCC(=O)N([C@H](C)CO)C[C@H]1C',
-]
-CLASSIFICATION_PROPERTIES = np.array([0, 1])
-REGRESSION_PROPERTIES = np.array([1.2, 2.3])
-
-def setup_test_directory():
-    """Create and return test directory path."""
-    test_dir = Path("test_model_checkpoints")
-    if test_dir.exists():
-        shutil.rmtree(test_dir)
-    test_dir.mkdir()
-    return test_dir
-
-def cleanup_test_directory(test_dir):
-    """Clean up test directory."""
-    if test_dir.exists():
-        shutil.rmtree(test_dir)
-
-def test_local_save_load():
-    """Test local save and load functionality for all models."""
-    print("\n=== Testing Local Save/Load ===")
-    
-    test_dir = setup_test_directory()
-    
-    models = {
-        'gnn': (GNNMolecularPredictor(num_task=1, task_type="classification"), CLASSIFICATION_PROPERTIES),
-        'grea': (GREAMolecularPredictor(num_task=1, task_type="classification"), CLASSIFICATION_PROPERTIES),
-        'sgir': (SGIRMolecularPredictor(num_task=1, task_type="regression"), REGRESSION_PROPERTIES),
-    }
-    
-    for model_name, (model, properties) in models.items():
-        print(f"\nTesting {model_name.upper()} model:")
-        
-        # Test save_to_local and load_from_local
-        save_path = test_dir / f"{model_name}_local.pt"
-        try:
-            # Fit and save model
-            model.fit(SMILES_LIST, properties, X_unlbl=SMILES_LIST)
-            model.save_to_local(str(save_path))
-            print(f"✓ Successfully saved model to {save_path}")
-            
-            # Load model
-            new_model = type(model)(num_task=1, task_type=model.task_type)
-            new_model.load_from_local(str(save_path))
-            print("✓ Successfully loaded model")
-            
-            # Verify loaded model works
-            predictions = new_model.predict(SMILES_LIST)
-            print("✓ Successfully made predictions with loaded model")
-            
-        except Exception as e:
-            print(f"✗ Error in local save/load test: {str(e)}")
-    
-    cleanup_test_directory(test_dir)
-
-def test_generic_save_load():
-    """Test generic save and load methods."""
-    print("\n=== Testing Generic Save/Load Interface ===")
-    
-    test_dir = setup_test_directory()
-    
-    # Test with GNN model as example
-    model = GNNMolecularPredictor(num_task=1, task_type="classification")
-    model.fit(SMILES_LIST, CLASSIFICATION_PROPERTIES)
-    
-    # Test local path
-    local_path = test_dir / "generic_local.pt"
-    try:
-        # Test save method
-        model.save(path=str(local_path))
-        print("✓ Successfully saved model using generic save")
-        
-        # Test load method
-        new_model = GNNMolecularPredictor(num_task=1, task_type="classification")
-        new_model.load(path=str(local_path))
-        print("✓ Successfully loaded model using generic load")
-        
-    except Exception as e:
-        print(f"✗ Error in generic save/load test: {str(e)}")
-    
-    cleanup_test_directory(test_dir)
-
-def test_huggingface_interface():
-    """Test Hugging Face interface with mocked upload/download."""
-    print("\n=== Testing Hugging Face Interface ===")
-    
-    test_dir = setup_test_directory()
-    
-    # Test with GREA model as example
-    model = GREAMolecularPredictor(
-        num_task=1,
-        task_type="regression",
-        gamma=0.8,
-        num_layer=2,
-        hidden_size=64,
-        batch_size=4,
-        epochs=2,
-        verbose=False,
-        model_name='GREA_TEST'
-    )
-    
-    # Fit model with sample data
-    model.fit(SMILES_LIST, REGRESSION_PROPERTIES)
-    
-    # Mock data for testing
-    repo_id = "liuganghuggingface/test-torch-molecule"
-    metrics = {
-        "mse": 0.123,
-        "r2": 0.456
-    }
-
-    # Test HF save interface
-    with patch('huggingface_hub.HfApi') as mock_hf_api, \
-         patch('torch.save') as mock_torch_save, \
-         patch('json.dump') as mock_json_dump, \
-         patch('builtins.open', create=True) as mock_open:
-        
-        # Mock HF API methods
-        mock_api_instance = mock_hf_api.return_value
-        mock_api_instance.create_repo = MagicMock()
-        mock_api_instance.upload_file = MagicMock()
-        
-        try:
-            # Test model upload
-            model.save_to_hf(
-                repo_id=repo_id,
-                metrics=metrics,
-                commit_message="Test upload",
-                private=False
-            )
-            
-            # Verify repo creation
-            mock_api_instance.create_repo.assert_called_once_with(
-                repo_id=repo_id,
-                private=False,
-                exist_ok=True
-            )
-            
-            # Verify model state was saved
-            mock_torch_save.assert_called()
-            
-            # Verify metadata and metrics were saved
-            mock_json_dump.assert_called()
-            
-            print("✓ Successfully tested model upload to HF")
-            
-        except Exception as e:
-            print(f"✗ Error in HF save test: {str(e)}")
-
-    # Test HF load interface
-    with patch('huggingface_hub.hf_hub_download') as mock_hf_download, \
-         patch('torch.load') as mock_torch_load, \
-         patch('json.load') as mock_json_load:
-        
-        # Mock the downloaded checkpoint
-        mock_checkpoint = {
-            "model_state_dict": MagicMock(),
-            "hyperparameters": {
-                "num_task": 1,
-                "task_type": "regression",
-                "gamma": 0.8,
-                "num_layer": 2,
-                "hidden_size": 64
-            },
-            "version": "1.0.0"
-        }
-        mock_torch_load.return_value = mock_checkpoint
-        
-        # Mock metadata and metrics
-        
-        try:
-            # Test model download
-            download_path = test_dir / "downloaded_model.pt"
-            mock_hf_download.return_value = str(download_path)
-            
-            new_model = GREAMolecularPredictor(
-                num_task=1,
-                task_type="regression"
-            )
-            new_model.load_from_hf(repo_id=repo_id, path=str(download_path))
-            
-            # Verify download was attempted
-            mock_hf_download.assert_called()
-            
-            # Verify model state was loaded
-            mock_torch_load.assert_called()
-            
-            print("✓ Successfully tested model download from HF")
-            
-        except Exception as e:
-            print(f"✗ Error in HF load test: {str(e)}")
-
-    # Test error cases
-    print("\n=== Testing HF Error Cases ===")
-    
-    # Test invalid repo_id
-    with patch('huggingface_hub.HfApi') as mock_hf_api:
-        mock_api_instance = mock_hf_api.return_value
-        mock_api_instance.create_repo.side_effect = Exception("Invalid repo")
-        
-        # try:
-        # model.save_to_hf(repo_id="invalid/repo")
-        # print("✗ Should have raised error for invalid repo")
-        # except Exception:
-        #     print("✓ Correctly handled invalid repo_id")
-    
-    # Test missing token
-    with patch('os.environ', {}):
-        # try:
-        # get HF_TOKEN from environment variable
-        # token = os.getenv("HUGGINGFACE_TOKEN")  # If set as an environment variable   )
-        model.save_to_hf(repo_id=repo_id, private=True)
-        print("✗ Should have raised error for missing token")
-
-        # except ValueError:
-        #     print("✓ Correctly handled missing HF token for private repo")
-    
-    # Test invalid model state
-    model.is_fitted_ = False
-    try:
-        model.save_to_hf(repo_id=repo_id)
-        print("✗ Should have raised error for unfitted model")
-    except ValueError:
-        print("✓ Correctly handled unfitted model upload attempt")
-    
-    cleanup_test_directory(test_dir)
-
-def test_supervised_encoder_save_load():
-    """Test save/load functionality for SupervisedMolecularEncoder."""
-    print("\n=== Testing Supervised Encoder Save/Load ===")
-    
-    test_dir = setup_test_directory()
-    
-    # Initialize encoder with predefined tasks
-    encoder = SupervisedMolecularEncoder(
-        predefined_task=["morgan", "maccs"],
-        epochs=2,
-        verbose=False
-    )
-    
-    try:
-        # Fit encoder
-        encoder.fit(SMILES_LIST)
-        
-        # Test local save/load
-        save_path = test_dir / "encoder_local.pt"
-        encoder.save_to_local(str(save_path))
-        print("✓ Successfully saved encoder")
-        
-        # Load encoder
-        new_encoder = SupervisedMolecularEncoder(
-            predefined_task=["morgan", "maccs"],
-            epochs=2
-        )
-        new_encoder.load_from_local(str(save_path))
-        print("✓ Successfully loaded encoder")
-        
-        # Verify loaded encoder works
-        encodings = new_encoder.encode(SMILES_LIST)
-        print("✓ Successfully generated encodings with loaded encoder")
-        print(f"  Encoding shape: {encodings.shape}")
-        
-    except Exception as e:
-        print(f"✗ Error in encoder save/load test: {str(e)}")
-    
-    cleanup_test_directory(test_dir)
-
-def test_error_handling():
-    """Test error handling in save/load operations."""
-    print("\n=== Testing Error Handling ===")
-    
-    model = GNNMolecularPredictor(num_task=1, task_type="classification")
-    
-    # Test loading non-existent file
-    try:
-        model.load_from_local("non_existent_file.pt")
-        print("✗ Should have raised FileNotFoundError")
-    except FileNotFoundError:
-        print("✓ Correctly handled non-existent file")
-    
-    # Test loading without fitting
-    try:
-        model.save_to_local("unfitted_model.pt")
-        print("✗ Should have raised error for unfitted model")
-    except Exception as e:
-        print("✓ Correctly handled unfitted model save attempt")
-    
-    # Test generic save without path or repo_id
-    try:
-        model.save()
-        print("✗ Should have raised ValueError")
-    except ValueError:
-        print("✓ Correctly handled missing path and repo_id")
-
-if __name__ == "__main__":
-    # Run all IO tests
-    test_local_save_load()
-    test_generic_save_load()
-    test_huggingface_interface()
-    test_supervised_encoder_save_load()
-    test_error_handling()