fixed model training

ANNOTATE/core/real_training_adapter.py

@@ -815,8 +815,8 @@ class RealTrainingAdapter:
                     continue
             
             if entry_index is None:
-                logger.warning(f"Could not find entry timestamp in market data")
+                logger.debug(f"Could not find entry timestamp in market data - using first candle as entry")
-                return negative_samples
+                entry_index = 0  # Use first candle if exact match not found
             
             # If exit not found, estimate it
             if exit_index is None:
@@ -1958,7 +1958,7 @@ class RealTrainingAdapter:
                                  'training_id': session.training_id
                              },
                             file_path=checkpoint_path,
-                            performance_score=float(avg_accuracy),  # Use accuracy as score
+                            file_size_mb=os.path.getsize(checkpoint_path) / (1024 * 1024) if os.path.exists(checkpoint_path) else 0.0,
                             is_active=True
                         )
                         

ANNOTATE/web/app.py

@@ -213,7 +213,7 @@ class AnnotationDashboard:
                         SELECT checkpoint_id, performance_metrics, timestamp, file_path
                         FROM checkpoint_metadata
                         WHERE model_name = ? AND is_active = TRUE
-                        ORDER BY performance_score DESC
+                        ORDER BY timestamp DESC
                         LIMIT 1
                     """, (model_name.lower(),))
                     

NN/models/advanced_transformer_trading.py

@@ -1229,7 +1229,7 @@ class TradingTransformerTrainer:
                     for k, v in batch.items()}
             
             # Use automatic mixed precision (FP16) for memory efficiency
-            with torch.cuda.amp.autocast(enabled=self.use_amp):
+            with torch.amp.autocast('cuda', enabled=self.use_amp):
                 # Forward pass with multi-timeframe data
                 outputs = self.model(
                     price_data_1s=batch.get('price_data_1s'),

NN/models/cob_rl_model.py

@@ -323,7 +323,7 @@ class COBRLModelInterface(ModelInterface):
         self.optimizer.zero_grad()
         
         if self.scaler:
-            with torch.cuda.amp.autocast():
+            with torch.amp.autocast('cuda'):
                 outputs = self.model(features)
                 loss = self._calculate_loss(outputs, targets)
             

NN/models/dqn_agent.py

@@ -1436,7 +1436,7 @@ class DQNAgent:
             import warnings
             with warnings.catch_warnings():
                 warnings.simplefilter("ignore", FutureWarning)
-                with torch.cuda.amp.autocast():
+                with torch.amp.autocast('cuda'):
                     # Get current Q values and predictions
                     current_q_values, current_extrema_pred, current_price_pred, hidden_features, current_advanced_pred = self._safe_cnn_forward(self.policy_net, states)
                     current_q_values = current_q_values.gather(1, actions.unsqueeze(1)).squeeze(1)

core/realtime_rl_cob_trader.py

@@ -536,7 +536,7 @@ class RealtimeRLCOBTrader:
             features_tensor = torch.from_numpy(features).unsqueeze(0).to(self.device)
             
             with torch.no_grad():
-                with torch.cuda.amp.autocast():
+                with torch.amp.autocast('cuda'):
                     outputs = model(features_tensor)
                     
                     # Extract predictions
@@ -934,7 +934,7 @@ class RealtimeRLCOBTrader:
             ], dtype=torch.float32).to(self.device)
 
             # Forward pass with mixed precision
-            with torch.cuda.amp.autocast():
+            with torch.amp.autocast('cuda'):
                 outputs = model(features)
 
                 # Calculate losses