AbstractIntegratedModule 1.1.4__tar.gz → 1.1.5__tar.gz

{abstractintegratedmodule-1.1.4 → abstractintegratedmodule-1.1.5}/AbstractIntegratedModule.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: AbstractIntegratedModule
-Version: 1.1.4
+Version: 1.1.5
 Summary: Integrated Pipeline with Specialized Non-LLM AI Agent Framework for ARM64 architecture
 Author: Micro-Novelty
 Author-email: Micro-Novelty <hernikpuspita5@gmail.com>

{abstractintegratedmodule-1.1.4 → abstractintegratedmodule-1.1.5}/AbstractIntegratedModule.py

@@ -4003,6 +4003,7 @@ class ExplainabilityModule:
     def data_preparation(self, titles, labels):
         datasets = []
         raw = []
+
         for title in titles:
             tupled_title = (str(title))
             datasets.append(tupled_title)
@@ -4025,6 +4026,7 @@ class ExplainabilityModule:
         filled = int(value * max_width)
         return '█' * filled + '░' * (max_width - filled)
 
+
     def _learn_from_feedback(self, text, correct_label, wrong_result, batch_size=2):
         eps = 1e-5
         print(f"\n[📚] Learning: '{text}' → {correct_label}...")
@@ -9872,6 +9874,7 @@ class IntegratedPipeline:
                 f"Unexpected memory type: {type(memory).__name__}",
                 "Clear and reinitialize memory")
 
+
     def k_fold_cross_validate(self, X, y, input_dim, n_classes, k=5, seed=42,
                             epochs=None, lr=None, log_to_diagnostics=True):
         """
@@ -13318,6 +13321,7 @@ class IntegratedPipeline:
             raise Warning('[!] Dataset is None or empty! Make sure you provide a dataset or create it automatically.')
 
         if not self.model2:
+            print(datasets)
             intents = [d[1] for d in datasets]
             intent_to_id = {intent: i for i, intent in enumerate(sorted(set(intents)))}
             num_classes = self._get_num_classes(label_map=label_map)
@@ -14505,6 +14509,7 @@ class AsyncResultQueue:
         logger.debug(f"[=] Submitted request {request_id}: {texts}")
         return request_id
     
+
     async def wait_for_result(self, request_id: str, timeout: int = 30) -> Dict:
         """
         Wait for a specific request to complete.
@@ -16369,7 +16374,6 @@ class PipelineAsyncManager:
 class PipelinePredictionManager:
     def __init__(self, pipeline, label_csv='labels.csv', target_title='title', label='label'):
         self.pipeline = pipeline
-
         try:
             print("📖 Loading labels from text file...")
             self.titles, self.y_raw, self.label_map = self.load_labels_from_csv(label_csv, target_title, label)
@@ -16490,8 +16494,10 @@ class PipelinePredictionManager:
             X_gen = None
             use_embedded = False
             attn_weights = None
+
             trans_probs = None
             mlp_probs = None
+            target_probs = None
 
             print(f"\n[🚀] Regular Prediction Initiated...")
             self.pipeline.titles = titles
@@ -16508,6 +16514,30 @@ class PipelinePredictionManager:
                     _, y, _, _ = self.pipeline.mlp_training_features(rules, dataset)                  
                 else:
                     dataset, _ = self.pipeline.data_preparation(titles, label_map)
+                    
+            if X is not None or y is not None and isinstance(X, (np.ndarray, list)) and isinstance(y, (np.ndarray, list)) and len(X) > 0 and len(y) > 0:
+                X_train, X_val, y_train, y_val = self._prepare_train_val_split(
+                    X, y, min_val_per_class=5, min_frac=0.1, max_frac=0.3
+                )
+                
+                onehot_validation = self.pipeline._validate_onehot(y_train)
+                if onehot_validation:
+                    if num_classes != len(label_map):
+                        num_classes = len(label_map)
+                    if num_classes > np.max(y_train):
+                        y_train = np.eye(num_classes)[np.asarray(y_train)]
+                        y_val = np.eye(num_classes)[np.asarray(y_val)]
+                    else:
+                        print('[⚠️] Warning: Y onehot encoding fails, Returning y samples as is, This may cause Exploding Gradient in MLP Training!')
+                        y_train = y_train
+                        y_val = y_val
+
+                X_mean = X_train.mean(axis=0)
+                X_std = X_train.std(axis=0) + 1e-8
+
+                X_train = (X_train - X_mean) / X_std
+                X = (X_val - X_mean) / X_std
+                y = y_val.copy()
 
             if X_gen is not None:
                 self.pipeline.transformer_utilities(X_provided=X, X_raw=X_gen, y_true=y, rules=rules, datasets=dataset, label_map=label_map, batch_size=batch_size)
@@ -16860,6 +16890,7 @@ class PipelinePredictionManager:
         self.pipeline.titles = titles
         self.pipeline.labels = label_map   
 
+        num_classes = self.pipeline._get_num_classes(label_map=label_map)
         try:
 
             if titles is not None and rules is not None:
@@ -16871,10 +16902,34 @@ class PipelinePredictionManager:
                 else:
                     datasets, _ = self.pipeline.data_preparation(titles, label_map)
 
+            if X is not None or y is not None and len(X) > 0 and len(y) > 0:
+                X_train, X_val, y_train, y_val = self._prepare_train_val_split(
+                    X, y, min_val_per_class=5, min_frac=0.1, max_frac=0.3
+                )
+                
+                onehot_validation = self.pipeline._validate_onehot(y_train)
+                if onehot_validation:
+                    if num_classes != len(label_map):
+                        num_classes = len(label_map)
+                    if num_classes > np.max(y_train):
+                        y_train = np.eye(num_classes)[np.asarray(y_train)]
+                        y_val = np.eye(num_classes)[np.asarray(y_val)]
+                    else:
+                        print('[⚠️] Warning: Y onehot encoding fails, Returning y samples as is, This may cause Exploding Gradient in MLP Training!')
+                        y_train = y_train
+                        y_val = y_val
+
+                X_mean = X_train.mean(axis=0)
+                X_std = X_train.std(axis=0) + 1e-8
+
+                X_train = (X_train - X_mean) / X_std
+                X = (X_val - X_mean) / X_std
+                y = y_val.copy()
+
             if X_gen is not None:
-                self.pipeline.transformer_utilities(X_provided=X, X_raw=X_gen, y_true=y, rules=rules, datasets=dataset, label_map=label_map, batch_size=batch_size)
+                self.pipeline.transformer_utilities(X_provided=X_train, X_raw=X_gen, y_true=y_train, rules=rules, datasets=dataset, label_map=label_map, batch_size=batch_size)
             else:
-                self.pipeline.transformer_utilities(X_provided=X, X_raw=X, y_true=y, rules=rules, datasets=dataset, label_map=label_map, batch_size=batch_size)
+                self.pipeline.transformer_utilities(X_provided=X_train, X_raw=X, y_true=y_train, rules=rules, datasets=dataset, label_map=label_map, batch_size=batch_size)
 
             reverse_map = {v: k for k, v in label_map.items()}
 
@@ -17395,7 +17450,7 @@ class PipelinePredictionManager:
             signals.append(f'AME={AME:.3f}')
 
         if error_counts is not None and len(error_counts) > 0:
-            # FIX — max, not mean: catches localized single-class failure
+            # max: catches localized single-class failure
             # that a flat average would dilute away
             worst_class_idx = int(np.argmax(error_counts))
             max_error = float(error_counts[worst_class_idx])
@@ -17434,6 +17489,7 @@ class PipelinePredictionManager:
             AME = None
             use_embedded = False
             dataset = None
+            target_probs = None
 
             X_gen = None
             sec_chosen_label = None
@@ -17661,6 +17717,7 @@ class PipelinePredictionManager:
                     return result, cached['prediction'], cached['confidence']
                 else:
                     print(f'[!] Similarity: {cached['similarity']} is low, Cannot pick label due to low certainty, Initiating advanced prediction...')
+                    target_probs = mlp_probs.copy()
             else:
                 print('[=] No verified output from cache available that matched samples, starting advanced prediction...')
                 if self.pipeline.use_transformer:
@@ -17669,7 +17726,7 @@ class PipelinePredictionManager:
 
                     target_probs = self.pipeline.predict_proba(input_ids, X, type='Hybrid', embedded=use_embedded)
                 else:
-                    target_probs = mlp_probs
+                    target_probs = mlp_probs.copy()
            
                 target_probs = target_probs[:mlp_probs.shape[0], :mlp_probs.shape[1]] 
                 target_probs = self.pipeline.model3.continuous_predictive_correction(self, target_probs, mlp_pred_indices)  
@@ -18025,7 +18082,6 @@ class PipelinePredictionManager:
                 if titles is not None and len(titles) > 0:
                     correct, sec_correct = self.display_hybrid_results(payload, final_class_idx, results, top_k, verbose=True)
                     
-                
                 return results, chosen_label, confidence
 
             else:
@@ -18312,7 +18368,7 @@ class PipelinePredictionManager:
             print(f'[⚡] Final Prediction: {chosen_label} with confidence: {confidence:.1%}')
             chosen_label = chosen_label
 
-        
+        self.pipeline.evaluate_mlp_performance(X, y, self.label_map)
         # delete pipelines cache
         print('[🔍] Pipelines Cache Cleaned!')
         self.pipeline.cache.clear()
@@ -18374,10 +18430,7 @@ class PipelinePredictionManager:
 
         except Exception as e:
             print(f'[!] Cant check and calibrate probs based on penalty due to: {e}') 
-            traceback.print_exc()  
 
-         
-        
         return final_probs