SIENTIAPDE-1253: Refactor training workflow and activities to raise exceptions on failure

This commit refactors the training workflow and associated activities to raise exceptions on failure instead of returning success/failure dictionaries. This allows the Temporal workflow to handle errors more effectively and ensures that the workflow stops when a critical error occurs.

Key changes:

- The train_model workflow is introduced to orchestrate the entire training process, including parameter validation, data download, model training, and model saving.
- The validate_train_params activity is added to validate and convert training parameters.
- The train_model and save_model activities are updated to raise exceptions on failure.
- The ExperimentStatus enum is updated to include a new status for orchestrator validation errors.
- The tests are updated to reflect the new exception-based error handling.
- The activities now return the TrainModelResult directly instead of a dictionary.

tests/activities/test_mlflow.py

@@ -1,6 +1,7 @@
 from unittest.mock import ANY, MagicMock, patch
 
 import numpy as np
+import pytest
 from pytest import fixture, mark
 from sientia_do.notifications.models import NotificationLevel
 from sientia_do.temporal.constants import DATETIME_FORMAT, DATETIME_FORMAT_WITH_TZ
@@ -333,10 +334,9 @@ async def test_save_model_success(mlflow):
     mlflow.model_monitoring_repository.generate_artifacts.assert_called_once_with(train_result)
     mlflow.model_monitoring_repository.save_run.assert_called_once_with(train_result)
 
-    # Verify response
-    assert response['success'] is True
-    assert response['result'] == train_result
-    assert response['error_message'] is None
+    # Verify response - now returns TrainModelResult directly
+    assert response == train_result
+    assert response.run_name == 'test_experiment-1'
     assert train_result.run_name == 'test_experiment-1'
 
 
@@ -362,13 +362,9 @@ async def test_save_model_get_next_run_name_error(mlflow):
         'train_result': train_result,
     }
 
-    # Call the method
-    response = await mlflow.save_model(input_data)
-
-    # Verify error handling
-    assert response['success'] is False
-    assert response['result'] is None
-    assert 'MLflow connection error' in response['error_message']
+    # Call the method - should raise exception
+    with pytest.raises(Exception, match='MLflow connection error'):
+        await mlflow.save_model(input_data)
 
     # Verify notification was sent
     mlflow.send_notification.assert_called_once_with(
@@ -404,13 +400,9 @@ async def test_save_model_generate_artifacts_error(mlflow):
         'train_result': train_result,
     }
 
-    # Call the method
-    response = await mlflow.save_model(input_data)
-
-    # Verify error handling
-    assert response['success'] is False
-    assert response['result'] is None
-    assert 'Reports directory does not exist' in response['error_message']
+    # Call the method - should raise exception
+    with pytest.raises(FileNotFoundError, match='Reports directory does not exist'):
+        await mlflow.save_model(input_data)
 
     # Verify notification was sent
     mlflow.send_notification.assert_called_once_with(
@@ -447,13 +439,9 @@ async def test_save_model_save_run_error(mlflow):
         'train_result': train_result,
     }
 
-    # Call the method
-    response = await mlflow.save_model(input_data)
-
-    # Verify error handling
-    assert response['success'] is False
-    assert response['result'] is None
-    assert 'One or more metrics (MSE, R2, MAE) are None' in response['error_message']
+    # Call the method - should raise exception
+    with pytest.raises(ValueError, match=r'One or more metrics \(MSE, R2, MAE\) are None'):
+        await mlflow.save_model(input_data)
 
     # Verify notification was sent
     mlflow.send_notification.assert_called_once_with(
@@ -491,10 +479,9 @@ async def test_save_model_missing_metadata(mlflow):
     # Call the method
     response = await mlflow.save_model(input_data)
 
-    # Verify it still works (metadata defaults to {})
-    assert response['success'] is True
-    assert response['result'] == train_result
-    assert response['error_message'] is None
+    # Verify it still works (metadata defaults to {}) - returns TrainModelResult directly
+    assert response == train_result
+    assert response.run_name == 'test_experiment-1'
 
 
 @mark.asyncio
@@ -540,10 +527,8 @@ async def test_save_model_complete_flow(mlflow):
     mlflow.model_monitoring_repository.generate_artifacts.assert_called_once()
     mlflow.model_monitoring_repository.save_run.assert_called_once_with(updated_result)
 
-    # Verify response
-    assert response['success'] is True
-    assert response['result'] == updated_result
-    assert response['error_message'] is None
-    assert updated_result.run_name == 'production_model-5'
-    assert updated_result.run_dir is not None
-    assert updated_result.report_path is not None
+    # Verify response - returns TrainModelResult directly
+    assert response == updated_result
+    assert response.run_name == 'production_model-5'
+    assert response.run_dir is not None
+    assert response.report_path is not None